Graduate School of Development Studies

A Research Paper presented by:

Nandini.Ramamurthy(India)

in partial fulfillment of the requirements for obtaining the degree of

MASTERS OF ARTS IN DEVELOPMENT STUDIES

Specialization:Economics of Development

(ECD: 09-10)

Members of the examining committee:

Supervisor’s name (Prof. Dr. Arjun Singh Bedi)Reader’s name (Prof. Howard Nicholas)

Internal Migration and Domestic Remittance: Study on Rural and Urban districts of Orissa and Gu-

jarat-India

The Hague, The NetherlandsNovember, 2010

ii

Disclaimer:This document represents part of the author’s study pro-gramme while at the Institute of Social Studies. The views stated therein are those of the author and not necessarily those of the Institute.

Inquiries:Postal address: Institute of Social Studies

P.O. Box 297762502 LT The HagueThe Netherlands

Location: Kortenaerkade 122518 AX The HagueThe Netherlands

Telephone: +31 70 426 0460Fax: +31 70 426 0799

iii

Acknowledgement:

My sincere thanks to Professor Arjun Singh Bedi for guiding and supporting me through out the research process. I am particularly grateful to my second reader for providing the critical comments that helped me understand my research from critical perspective. My heartfelt thanks to my mother for keeping constant faith in me to do this research work and for providing all the necessary inputs for making me possible to collect such a large volume of data. My sincere appreciations to Swati for making me think a lot on my questionnaire and the support extended by her during that period. I extend my thanks to my peer group readers for their valuable feedback. I am deeply indebted to those 11 enumerators and respondents who made it possible to collect large amount of data in a short period of time. Special thanks to Anil, Naresh and Sreenu for working upto three in the morning for collecting the data and very patiently listening to whatever I had to say during the field process. Finally, Special gratitude to my family members and especially to Prasanna with whom I could share all my experience and making my field study even more eventful.

iv

ContentsList of Tables viList of Figures viiList of Acronyms viiiNSS- National Sample Survey viiiNSSO- National Sample Survey Organisation viiiNCRL- National Commission on Rural Labour viiiGDP- Gross Domestic Product viiiIFMR- Institute of Financial Management Research viiiNGO- Non Governmental Organisation viiiLPM- Linear Probability Model viiiOLS- Ordinary Least Square viiiDFID- Department for International Development viiiWORLP- Western Orissa Rural Livelihood Project viiiST- Schedule Tribe viiiSC- Schedule Caste viiiAbstract xAbstract x

Chapter 1-Introduction 11.1 Background 4

The context and Relevance of the Study 4

Chapter 2 - Field Work 62.1 Methodology, Sampling procedure and Question-

naire 62.1.1 Methodology 62.1.2 Sampling procedure (Orissa and Gujarat) 7

2.1.2.1 Data collection procedure (Orissa and Gu-jarat) 7

2.1.3 Questionnaire 8

Chapter 3 – Theoretical and Empirical Results 103.1 Factors driving migration and socio-economic the

conditions of migrants 103.1.1 Theoretical Framework 10

Classical theory of migration: 11The role of agriculture in migration: 12The role of Industrialization and Urbanisation in mi-

gration: 15

v

The role of excessive debt burdens: 163.1.2 Empirical Analysis 17

Model specification: 18Variable Specification 18Rational choice for Probit Model: 19

3.1.3 Empirical Results 20a. Factors driving migration and migrant profile 20b. Living and working conditions of migrants 26c. Migrant-illness to living and work conditions 30

Chapter 4-Discussion and Empirical Analysis on Cost of Remittance 364.1 Need and potential effect of remittances: 363.2 Money Transfer Mechanism 383.3 Empirical Analysis: 40

Empirical Results 40

Chapter 5 - Conclusion 47Appendices 49References 64

vi

List of TablesTable 1 Survey Questionnaire Outline collected from

Orissa and Gujarat 8Table2 Share of Agriculture in GDP and Employment 13Table 3 Commodities Group-wise Value of Trade (Rs in

Lakh Crores) 13Table 4: Motivation for Migration: Extended Household20Table 5: Age distribution of migrants 21Table 6: Distribution of Sex, Caste and Religion of mi-

grant: 22Table 7: Distribution of State, District and City of mi-

grant: 23Table 8: Education level of migrant 24Table 9: Details on Marital Status, Family Migration and

Child Migration 25Table 10: Migrant occupation in informal sector at Surat

and Gandhidham 26Table 11: Migrant working hours 27Table 12: Wages before and after migration (last one

year) per annum 27Table 13: Living Arrangement of migrants according to

city and room rent (last one year) per annum 28Table 14: Living arrangement of migrant according to

city and room sharing basis 29Table 15: Availability of facilities at work place 30Table16: Availability of facilities at Living place 30Table 17: Health Problems faced by migrants (last one

year) 31Table 18: Multivariate Regression on migrant illness 32Table 19: Average Per capita income of migrants’ house-

holds at Origin by income source 34Table 20: Analysis of cost, frequency and size of remit-

tance depending on the frequency of remittance 40Table 21: amount, channels and cost of remittance

(these figures are based on per annum) 41Chart3: Size of remittance and channel used for remit-

tance 42

vii

Table 22: Cost of remittance with mean and plus/minus standard deviations 43

Table 23: Cost of remittance as a proportion of amount for each channel (per month) 44

Table 24: Information on the charges collected by M-PESA 44

Table 25: Average size of remittance to cost amount in Rupees (last one year) 45

Table 26: Percentage of income from various sources (per annum) 45

List of FiguresChart 1: Possession of land to number of migrants 14Chart 2: Period wise development of Agriculture to Non-

agriculture (per annum) 15

viii

List of Acronyms

NSS- National Sample Survey

NSSO- National Sample Survey Organisation

NCRL- National Commission on Rural Labour

GDP- Gross Domestic Product

IFMR- Institute of Financial Management Research

NGO- Non Governmental Organisation

LPM- Linear Probability Model

OLS- Ordinary Least Square

d-probit- Marginal effect of Probit DFID- Department for International Development

WORLP- Western Orissa Rural Livelihood Project

ST- Schedule Tribe

SC- Schedule Caste

BC- Backward class

GC- General Caste

FSP- Financial Service Providers

NELM- New Economics of Labour Migration

RBI- Reserve Bank of India

KYC- Know Your Customer

ix

hh- household

HH- Head of Household

UPR- Usual Place of Residence

M

x

Abstract

This paper broadly deals with internal migration and do-mestic remittance in India. The paper discusses on the role of migration and remittance from the perspective of remit-tance sending household. In case of migration it has been analysed that migrant illnesses do have a negative impact on migrant’s present and future working ability. Further, the paper discusses on cost of remittance, channels and size of remittance used by the migrant to remit money back home and the constraints faced by them while remitting the money. Finally, the importance of remittance to show that remittance is a crucial source of income for households liv-ing in the source place. The paper is based on a sample of 419 households in Orissa, comprising of 230, remittance re-ceiving households and 189, non-migrant households. In Gujarat 288, extended households/individuals information was collected. This adds to a total of 707, sample house-holds consisting of 2810 individuals.

Relevance to Development StudiesTypically the previous studies of migration and remit-

tance have been on the effect of remittances on poverty/other outcomes at the receiving end while ignoring con-sequences of migration from the perspective of the sender. In a broader sense existing papers do not capture the ex-tent and magnitude of socio economic conditions at the mi-grated end. In my research study specifically I look at in-ternal migration and domestic remittance in India from the perspective of remitting sending household/individual. Hence placed with in this context, it is worthwhile to under-take a study to understand the impact if any, does this mi-gration have on remittance sending household/individual on their present and future working abilities?

KeywordsInternal migration, domestic remittance, urban labour mar-kets, Orissa, Gujarat, remittance sending/remittance receiv-ing household

xi

xii

Chapter 1-Introduction

Movement is an integral part of human existence. A large number of movements like commuting to and from a place of work, travelling for business or pleasure are of a tempor-ary nature. While some movements are relatively perman-ent and involve a change of residence from one place to an-other, this is treated as migration. Though social, cultural, political, personal and natural forces have different kinds of bearing on migration, when viewed as an economic phe-nomenon it receives special attention (Migration in India, 2007-08, NSS 1 64TH round July 2007-June, 2008). The pro-cess of migration is not simple; it varies from being perman-ent, semi-permanent and temporary 2. It can also be viewed as voluntary or forced, legal or illegal (Sansristri 2008).

Migration, be it domestic or international, has univer-sally been considered as an alternate strategy of susten-ance for livelihood by a large number of poor families. Whether it is due to push factors (like agricultural failure, lack of employment and high debts) or pull factors (such as better wages and availability of jobs), poor people migrate to booming and intermediate cities, as well as manufactur-ing and industrial centres in search of employment. These migrants not only attempt to improve their own livelihoods, but also remit a considerable share of their earnings back to their families (Sahu and Das 2008).

In the countryside remittance still remains a major source of income for families. Such money is often earned through low wages, irregularity of income and employment, frequent retrenchments and pronounced absence of social security. The living environment for these migrants is sub-standard and unhygienic. Even though, they are trapped between the two worlds, a difficult earning and living envir-onment in cities, these workers continue to negotiate with the uncertainties terrains of the urban labour markets so that they can provide the crucial monetary support required for sustenance (Sahu and Das 2008).

In such conditions, the workers want and need to send their hard earned money in a fast and safe mode. They want an even quicker mode of acknowledgement of receipt, and its associated cost to be low (Sen 2007). Hence, providing these services fills an important gap in the provision of fin-ancial services to the poor, enabling them to increase ac-cess, decrease cost and facilitate the remittance of smaller

1

amounts more frequently to provide crucial income support to families left behind (Sahu and Das 2008).

1 NSS- National Sample Survey.

2Where people move for few months in a year, the migrants are also known as short duration migrants, seasonal migrants or cir-culatory migrants.

When migration and remittances are treated as a pro-motion of a livelihood strategy it can have positive, negative or neutral effects on the welfare of rural household and communities, depending on the type of household/com-munity. The impact also changes with time: at the begin-ning, migration may deprive the household and rural eco-nomy of labour but in the long term, when remittances are invested it improves productivity, it creates assets and gen-erates income at the household level (Deshingkar and Start 2003).

The migrant labour market makes enormous contribu-tion to the Indian economy 3 despite being poorly endowed. They come from poor families where access to physical, fin-ancial and human capital is limited and where prospects for improving living standards are constrained by their inferior social and political status (Deshingkar and Akter 2009).

Highlighting the importance of migration and its contri-bution to the Indian Economy it is worthwhile to under-stand the definition provided by National Sample Survey Organization of India (NSSO) in its 55th round conducted in months of June-July 1999-00, surveyed migration according to ‘last usual place of residence’ (UPR) and ‘birth place’. Data on migration is based on particulars of the sample households. The definition of migration given by NSSO, of India as follows:

“A member of a sample household is treated as a migrant, if he/she has stayed continuously for at least six months or more in a place (village/town) other than the village/town where he/she has been enumerated. The village/town where the person has stayed continuously for at least six months or more prior to moving to the place of enumera-tion (village/town) is referred to as the ‘last usual place of residence’ of that migrated person. Shifting of residence

2

within village/town is not considered as an event of migra-tion”.

Considering the above definition, The National Commis-sion on Rural Labour (NCRL 2001) puts the number of cir-cular migrants in rural areas at around 10 million (includ-ing roughly 4.5 million inter-state migrants and 6 million in-tra-State migrants). Official statistics indicate that internal migration has remained remarkably low, because these sur-veys focus on permanent migration and disguise substantial seasonal and temporary flows of labour.

It is topical to note that both the National Census and the National Sample Survey use definitions of migration that are not employment related but are based on birth-place 4 and change in last usual place of residence 5. Secondly, they count migrant stock and not the temporary flow which is actually more important. Lastly, they tend to underestimate short term movements and altogether over-look seasonal and circular migration which accounts for the bulk of migratory movements for work (Deshingkar and Ak-ter 2009). 3Migrant contribute 10% of India Gross domestic product (GDP), through major sectors such as construction, textile, small indus-tries, brick-making, stone quarries, mines, fish and prawn pro-cessing and hospitality services.4Migrant by place of birth are those who are enumerated in a vil-lage/town, other than their place of birth, at the time of census.5A person is considered a migrant by place of last residence if the place in which he is enumerated during the census is other than his place of immediate last residence. Capturing the latest of mi-gration would give a better picture of the current migration scen-ario.

Thus, in the Indian context almost no systematic data has been collected on the prevalence of these temporary la-bour movements, the conditions under which people mi-grate, the costs and risks of migration, or the impact of re-mittances on the household in the sending areas (IFMR 2009). With this short coming in the definition and underes-timation of official statistics it is not possible to capture the extent and magnitude the migration process or migratory processes. Hence, the findings of the census of India and NSS ought to be reviewed against this backdrop.

In order to gain a better understanding of migration one needs to understand the relationship between domestic mi-gration, return and development. Two questions that have attracted the attention of migration scholars and policy makers are (i) how does migration impact development and (ii) how is migration influenced by development. To address

3

these questions, attention has focused on determinants of migration and consequences in the areas of origin and des-tination as well as on the volume, patterns and dynamics of migration (Ammassari and Black 2001).

As mentioned earlier as there are different types of mi-gration, there are also varying types of return6. These dif-ferences are important factors influencing the impact of mi-gration, return migration7 and development (Ammassari and Black 2001). In this context the emphasis is on perman-ent return rather then temporary return8. In the entire mi-gration chain the emphasis is on positive consequence of financial flows in the source area. However, it neglects vari-ous types of mental, physical and social challenges which a migrant undergoes at the destination (Sahu and Das 2008).

Hence, the consequences of migration vary depending on important factors like socio-economic and cultural situ-ations as well as multiple other factors that influence the impact of migration on individuals and their families.

6 Here types of return refer to as temporary return and perman-ent return in the process of migration which depends on the type of migration. 7 Return migration is a phenomenon in which the migrants return to their earlier usual place of residence.8 If the nature of return is temporary then it is broadly determ-ined by the seasonal character of the activities available at the destination place. When they return their labour is used for cul-tivation purpose until they decide to return back to destination place.

Motivated by the lack of research on migrant remitting individuals, the main objectives of this study are to address the combined effect of migration and remittances at both ends of the migration chain, from the perspective of the mi-grating and remitting household member/s as well as from the perspective of the receiving and remittance receiving

4

household. The first research objective of this paper is to analyse factors that cause migration, examine the socio eco-nomic conditions of remittance sending individual, analyse factors that influence migrant illness and future working ability of the migrant. The second research objective is to examine the cost, channels, frequency and amount through which migrants remit money to their families and import-ance of remittance in the rural household.

1.1 Background

The context and Relevance of the StudyThe paper is based on internal migration and domestic re-mittance between rural and urban areas, the study com-prises 707 households in Orissa and Gujarat, situated in East and West part of India respectively. Orissa, for more than half a century has undergone frequent natural calamit-ies. There has been a substantial reduction in the availabil-ity of forest produce and a lack of employment opportunit-ies. Which resulted in heavy indebtness amongst people, and the absence of a robust economic infrastructure has pushed the Oriya people to migrate in various parts of India (Sen 2007). The proportion of people that live below the poverty line is estimated at 47.13% when compared to 26.1% for the country as a whole (Planning commission of India, 2001).

The above mentioned reasons have caused out-migra-tion that has taken many forms and patterns – seasonal/dis-tress and rural to urban (Samal 2006). The three study dis-tricts, Kurdha, Nayagad and Puri, have a long tradition of sending migrants to economically important pockets of Mumbai, Kolkata and Gujarat (Sahu and Das 2008). The state of Gujarat has been a major destination for Oriya mi-grants. In Surat the power loom and diamond industry are two of the core sectors that determine its economic vi-brancy or lack of it. In Gandhidham migrants mainly remain engaged in the jobs associated with Knitting of readymade garments, while at the port of Kandla they work as casual loaders and unloaders (Sahu and Das 2008). It is significant to note that a large number of Oriya migrants live in dingy conditions often in groups of 10 to 12. They work as well as sleep in shifts involving multiple occupancies in small rooms for specific spans in a day/night8. Indeed, one often does not even get three square feet of space to sleep.

5

Moreover, this has serious implication on their physical, emotional and psychological health (Sahu and Das 2008).8Occasionally, migrant attempts at reducing the rent space hence, they organize their sleeping behaviour in postures that compel them to keep their bodies in vertical positions or their sides rather than in a horizontal or flat position. The latter con-sumes a higher amount of space, resulting in larger amounts of rent shared equally by occupants of a unit. Such shift sleeping further burdens the inmates and disturbs the arrangement when a worker becomes ill and cannot vacate the space for his counter-part (Sahu and Das 2008).

Previous studies (Kothari 2002, Rele 1969 and Sahu and Das 2008) of migration and remittance have been on the effect of remittances on poverty/other outcomes at the re-ceiving end while ignoring consequences of migration from the perspective of the sender. In a broader sense existing papers do not capture the extent and magnitude of socio economic conditions at the migrated end. In my research study, I look at internal migration and domestic remittance in India from the perspective of the remitting sending households/individuals. Hence, placed with-in this context, it is worthwhile to undertake a study to understand the im-pact, if any, that migration has on remittance sending households/individuals on their present and future working abilities.

6

Chapter 2 - Field Work

2.1 Methodology, Sampling procedure and QuestionnaireTo meet the research objectives, a household survey was designed and carried out in the months of July and August 2010. The selected states were Orissa and Gujarat and the selected districts were Kurdha, Nayagad, Puri (Orissa), Surat and Kutch (Gujarat) respectively. The major focus the survey was to gather information on the effect of migration and remittance and the socio economic status of remittance receiving, non migrant’s and remittance sending house-holds. To address the various issues I designed two sets of questionnaires. Each set of questionnaires was broadly clas-sified into six sections. The details of these sections are lis-ted in the following sub-section. Prior to discussing the es-timation approach and research findings, the following sec-tion deals with methodology, sampling procedure and the questionnaire used to gather primary data for this study.

2.1.1 MethodologyThe field study employed both quantitative and qualitative methods that used collection instruments such as a house-hold survey, discussion with migrants and non-migrants and with officials of Non Government Organisation, NGO (Adhi-kar, Orissa). The study was divided into three phases begin-ning with a survey of revelant literature. This was followed by a pre-study survey and the final phase was an intensive field study. The main aim of these discussions was to strengthen understanding on the extent and magnitude of migration, from the perspective of remitting household/indi-vidual and from the perspective of the remittance receiving household. The concurrent research design enabled me to gain insight not found in the quantitative and qualitative part of migration. Based on revelant survey literature (Sahu and Das 2008, Misra 1998 and Planning commission of In-dia 2001), Orissa9 and Gujarat10 were selected for data col-lection.

The sampling was done in both urban and rural areas; in the case of a rural area, the stratification was done on the basis of districts, villages and hh. In case of an urban area, the stratification was done on the basis of city, area and hh. In the following section I discuss the sampling pro-cedure adapted at each place.

7

9 The planning commission of India (2001) show that Orissa’s be-low poverty line is around 39.8%, which was the highest when compared to rest of India, followed by Bihar (32.5%) and Madhya Pradesh (32.4%)10 The major work available to Oriya migrants is in the powerloom and diamond polishing sector of Surat. In Gandhidham they work as loaders and unloaders. The other prominent factories include plastic, fertilizers, salt manufacturing, pharmaceuticals and brick manufacturing. Most of these Oriya workers get job year round and this forms the choice for selecting cities like Surat and Gandhidham.

2.1.2 Sampling procedure (Orissa and Gujarat) The selection of these districts was based on a review of the literature (Sahu and Das 2008 and Misra 1998), which showed that in the beginning of the early eighties large numbers of younger migrants were from Ganjam districts, and the later migrants mainly came from Kurdha, Nayagarh and Puri districts. It was evident that these districts had a long tradition of migration, which was an important liveli-hood option. Therefore, I chose to work in the above men-tioned three districts.

In Orissa, the selection of these villages was based on purposive sampling. Initially I selected those villages that had more than 150 hh. The criteria for selection depended on the distribution of migrants and non-migrants in those villages, and whether the villages had migrants whom mi-grated to Surat and Gandhidham. Any village that had more than 50 migrated and 50 non-migrated hh were selected as potential village for survey. This was done to get a repres-entative sample for the total population.

The selection of households in these villages was based on snowball sampling. This sampling technique was adap-ted to gain access to those migrants and non-migrants that were distributed in the villages. We approached a villager which could help us identify one migrant hh and from there, based on acquaintances, we collected information on both migrant and non-migrant.

In the case of Gujarat, the selection of the districts was based on the review of literature (Sahu and Das 2008), which shows that the majority of migrants from the source districts migrated either to Surat or Kutch (Gandhidham). The selection of areas was based on purposive sampling. A

8

discussion with the NGO brought out that there were only certain areas where we could find people from these dis-tricts. Therefore those areas were selected for our sample collection. The selection of the hh was based on snowball sampling. The sampling was done on the mutual association of the NGO with migrants, as well as mutual relationships between migrants.

The paper is based on a sample of 419 households in Orissa, comprising of 230 remittance receiving households and 189 non-migrant households. In Gujarat 288 extended households/individuals information was collected. There were 707 sample households consisting of 2810 individuals. For pragmatic reasons a sampling technique used which builds up large enough data to perform a useful research and statistical analysis.

2.1.2.1 Data collection procedure (Orissa and Gujarat)

For the purpose of this study the researcher received the assistance of 7 enumerators at the source places. The re-searcher and enumerators were unaware of the distribution of migrants, within the research districts in the state of Orissa. Therefore it took the assistance of local people to identify the distribution of migrants in villages. When the team reached a village one of the experienced enumerator would collect information pertaining to the distribution of migrants and non-migrants as well as place of migration. If the village consisted of migrant and non-migrant house-holds exceeding 50 then two enumerators were allotted per village to collect data pertaining to migrants as well as non migrants. In this approach each team covered at least 4 to 6 villages per day11.

In order to generate primary data to understand the ex-tent of remittance flow and its transfer process, as well as for assessing levels of involvement of formal and informal channels in transferring remittances, the researcher selec-ted those migrants who have migrated to Surat and Gandhidham at least one year before. This is due to the fact that migrants are likely to remit money after one or two months of his arrival at the destination point.

Since there were only two enumerators for Oriya we di-vided the group into two, each group had one enumerator. Since most of the workers were working on a shift basis the enumerators went to work place to make appointments for interviews. In each area the team members had to convince the elder migrant by explaining the purpose of the study,

9

and more importantly there was no government interven-tion. Once they were convinced they were ready to share the information.

The respondents 12 from Surat and Gandhidham were very patient and spent a lot of time with our enumerators even after their strenuous work schedule. It took the team about 30 days to collect the 707 samples.

Initially, the research was designed to develop a link between households in Orissa and the family member/ members of that household who have migrated to Gujarat. Due to volatility in the nature of job and time constraints the research was only able to link 5% of the total sample collected to the 36 individuals at the destination point.

2.1.3 QuestionnaireThe questionnaire used during the field survey was compre-hensive in the sense that it had the following sections of in-formation.

Table 1 Survey Questionnaire Outline collected from Orissa and Gujarat

Orissa GujaratReceiving Household level socio economic condi-tions sample survey

Sending Household level socio economic condi-tions sample survey

Section I - General information Section I - General informationSection II – Household Roster Information Section II – Socio Economic conditions of mi-

grants Section III – Socio Economic Conditions of HH Section III – Occupational health hazards Section IV – Employment and income details Section IV – Employment and income detailsSection V – Information on remittance (cost, fre-quency, amount and channels)

Section V - Information on remittance (cost, fre-quency, amount and channels)

Section VI – Qualitative analysis on effect of mi-gration and remittance

Section VI – Qualitative analysis on effect of mi-gration and remittance

The questionnaire adopted included quantitative as well as qualitative aspects to capture information pertaining to personal details, family composition, ownership of land and other resources, source of income at native residence, their occupation (present as well as past), income, living condi-tions, size of remittance, earmark of remittance, savings be-

10

haviour and their perceptions regarding what the future holds for them.

11 When the team visited the 4 to 6 villages, if it was found that the distribution of migrant and non-migrant was unequal in these villages or that villages had migrants that had chosen to migrate to other places like Mumbai, Vapi, Tamil Nadu, Andhra Pradesh and Jammu & Kashmir, data was not collected as these places were not included in the study.2 To identify the respondents in the location of Surat, team mem-bers visited salon shop (hair-cut shop) and also received help from a local NGO. During the interaction we were able to estab-lish a link from Surat and Gandhidham by recording their mobile numbers of the people who have migrated from Surat.

11

Chapter 3 – Theoretical and Empirical Results For each of the stated research objectives this paper provides a theoretical framework, followed by empirical analysis. The paper deals with two issues (i) factors driving migration from the perspective of remittance sending indi-viduals and to assess their socio-economic conditions, and (ii) to examine the cost of remittance, amount, frequency and channels through which they remit their money back home and its impact on rural households. The objectives are analyzed within the Indian context emphasizing internal mi-gration from rural-to-urban areas. Accordingly, this paper has been broadly divided into two chapters to deal with the above stated research objectives.

3.1 Factors driving migration and socio-economic the conditions of migrantsIn this part I discuss factors that drive internal migration from rural-to-urban areas, the factors that affect migrant’s present and future working ability and migrant-illness. To address these issues I examine the theory of internal migra-tion and in particular push and pull factors that cause in-ternal migration in India. Thereafter I build a profile of these migrants and assess socio-economic conditions at the destination point.

3.1.1 Theoretical Framework In recent times the Indian population has been effected by factors influencing mobility patterns, such as the rapid transformation of the Indian economy, improvement in the levels of education, transportation and communication facil-ities, a shift in workforce from agriculture to industry and tertiary activities (B. Bhagat 2009). In contrast migration is sometimes thought of a promotion of a livelihood strategy intended for economic mobility among the poor. Such move-ments have taken place in search of work, in response to environmental shocks and stress and due to political con-flict. The growing migration rates13 are attributed to over-all development mainly due to the structural transformation in 1990s which has propelled growth and was followed by

12

the regional imbalance within the country. Apart from structural transformation migration may be understood through rural push and urban pull factors playing a role in understanding increased mobility between the origin and destination point.

13 As per (Das 1993) study, as high as 80% of the slum dwellers in the city have come from regions outside Surat and did not origin-ally belong to Surat. The increase is also attributed to increase utilization of the powerloom from 8105 in 1960 to 19025 in 1970, a growth rate of nearly 14% per year... The growth rate of looms was 21% in the seventies and 33% in the 80’s and in 1994; there are an estimated 2.5 lakh looms (Das 2007)

In this chapter I provide a broad overview of the theor-ies of migration using Harris-Todaro14 (HT) and Lewis model15 to assess the cause of internal migration from rural-urban areas. I use factors such as agriculture, industrialisa-tion, urbanization and debt burden.

Classical theory of migration:In the HT model migration is regarded as an adjustment to the mechanism by which workers allocate themselves between different labour markets, some of them are located in urban areas and some in rural areas, while attempting to maximize their expected incomes. Potential migrants re-spond to the probability of obtaining urban employment. The HT model treats rural-urban migration primarily as an economic phenomenon and demonstrates that, in certain parametric ranges, an increase in urban employment may actually result in higher levels of urban unemployment.

Thus, this model considers the formal and informal sec-tors that observe differentials in labour markets conditions in terms of minimum wage laws, pension scheme, unem-ployment benefits and other facilities that are required by law. Therefore, migration, in this model is viewed as a re-sponse to the significant wage gap that prevails between the two sectors. Since not everyone is observed in the formal sector. A consequence of rural to urban migration is the use of the urban informal sector as a source of employ-ment.

The Lewis model tells us that agricultural surpluses and labour must be transferred in tandem for industrial devel-opment to begin. The Lewis model provides an explicit ex-planation of how structural change, propelled by labour

13

transfer can bring about economic growth in a dual eco-nomy (traditional agricultural sector and the modern indus-trial sector). However, there are some crucial points that need to be considered when it comes to labour migration and unemployment, such as capacity of the industrial sector to absorb the surplus labour force of the rural sector which crucially depends on the amount of reinvested profit, the pace of capital accumulation and capital investment in tech-nologies to speed up the production procedures.

Hence according to Classical theories discussed above, migration is a rational decision made by an individual to move from a less advantageous situation, to a more advant-ageous one, after weighing risks and benefits. Thus consid-ering the two theories it is worth while to discuss internal migration from the position of push and pull factors.

14 Riadh 1998, Rural-Urban Migration: On the Harris-Todaro Model, University de Bretagne Sud, pp 1-28Fields 1974, Rural-Urban Migration, Urban Unemployment and Underemployment, and job-search activity in LDCs, Yale Univer-sity, pp 1-2315http://www.economywatch.com/unemployment/labor-mi-gration.html

The push and pull factors have dominated much of the understanding of migration. Low income, low literacy, de-pendence on agriculture and high poverty are some ex-amples of push factors associated with place of origin. On the other hand, high income, high literacy, dominance of in-dustries and services, are the pull factors associated with the destination. This suggests that higher income and the transformation of an economy from reliance on agriculture to the non-agriculture sector may translate into high levels of migration from rural to urban areas (Bhagat, 2009, Bhalla and Hazell, 2003 and Unni, 1998).

Migration of labour among the poor has a peculiar char-acteristic. It can be involuntary16, in the sense that the pro-spective migrants, in most cases make the decision to mi-grate based on their expectations of estimated gains from the movement. In some cases, the same movement could be considered as forced migration17 where often poverty, land-lessness, debt, unemployment act as the push factors for mass exodus (Chatterjee, 2006).

14

In order to gain an in-depth understanding of the push factors it is important to address issues relating to agricul-tural failure, lack of productivity, decline in price, landless-ness and absence of other livelihood opportunities, all of which generate mobility in search of better livelihood op-portunities.

The role of agriculture in migration:Stagnation and volatility of the agrarian economy and a lack of diversification within the sector has forced the poorer rural sections of the population to move in search of work and to join the lower circuits of the urban labour mar-ket. Migration for survival, documented by (Murthy, 1991; Reddy 1990; Rao, 1994) suggests that the main drivers of migration are the worsening situation of dry-land agricul-ture created by drought, crop failure and poor terms of trade.

A study by Bhalla and Hazell (2003) reveals that over the past three decades, the employment status in rural In-dia has undergone some important changes due to a drop in the price levels of agricultural products, in relation to Gross Domestic Product (GDP) that has led to movement of labour from rural to urban areas. Although agriculture continues to occupy a predominant position in the Indian economy, the relative price share of agriculture in relation to GDP has declined from 44.8 per cent in 1977-78 to only 27.6 per cent in 1999-2000 in comparison the price in 1993-94 was about 33.5 per cent. On the contrary, the share of employ-ment in agriculture declined from 73.9 per cent in 1972-73 to 60.2 per cent by 1999-2000. With 60 per cent of the na-tional workforce producing a little more than a 25 per cent of GDP, the relative productivity of workers in agriculture is less than one- fourth that of in non-agricultural occupations.

16 Involuntary migration is a result of extreme economic and of-ten social hardship, and is undertaken mostly by landless or land-poor, unskilled and illiterate poor labourers or people displaced by socio-political crisis. Here, there is no choice of the place or type of work they undertake17 Forced migrations refers to the involuntary movement of a per-son wishing to escape from armed conflicts, or a situation of viol-ence and/or the violations of his/her rights or a natural or man made disaster.

Table2 Share of Agriculture in GDP and Employment

Year Percent share of agriculture in Percent share of agriculture in

15

GDP at 1993-94 Prices Employment (UPSS) 1972-73 44.8 73.91993-94 33.5 63.91999-2000 27.6 60.2Source: National Accounts Statistics and Survey on Employment and Unemployment-various rounds

Table 3 Commodities Group-wise Value of Trade (Rs in Lakh Crores)

Commodity Groups

2004-2005 2005-2006 2006-2007 2007-2008

Bullion and other metals

1.80(31.47)

7.79(36.15)

21.29(57.10)

26.24(64.55)

Agriculture 3.90(68.18)

11.92(55.31)

13.17(35.82)

9.41(23.15)

Energy 0.02(0.35)

1.82 (8.45)

2.31 (6.28)

5.00(12.30)

Others 0.00(0.00)

0.02(0.09)

0.001(0.00)

0.00(0.00)

Total 5.72(100.00)

21.55(100.00)

36.77(100.00)

40.65(100.00)

Source: Ministry of Consumer Affairs, Food & Public Distribution Government of India, 2008Note: Figures in parenthesis indicate percentage to total value

From the above analysis it is clear that even though the rate of employment in the agriculture sector has not de-creased drastically, its productivity relative to non-farming sector has reduced. From the above table it is evident that the fall in the price of agricultural commodities over a period of time has led to low productivity in the primary sector of the rural economy. These reasons suggest that a withdrawal of participation due to a lack of economic in-centive has forced them to take up non-farming activities that are readily available in the urban economy.

Another study by Jayati and Chandrasekhar (2003) re-veals that, among 13 states in India a significant proportion of the people were found to be practicing in part-time farm-ing, which is about 35% in the year 1987-88.This proportion increased to 41% in 1999-2000. It is estimated that 41% of the youth migrated for work outside the village and most of this migration was seasonal. The key point for such a pulling out from agriculture was the absence of irrigation and lack of economic incentive which led to rural-urban mi-gration.

In addition to falling productivity and a lack of irrigation in some states, landlessness is a factor motivating migra-tion. As displayed below, according to NSSO in its 49th

round surveyed in the months of January-June 1993, land is an important indicator and also a sign of affluence. If a household has little land holding then the likelihood of mi-gration increases. The graph below shows that migration amongst households possessing no/very little land is about

16

four times higher than households possessing more than 8.0 hectare of land

Chart 1: Possession of land to number of migrants

Source: NSSO 49TH round (January-June 1993) nationwide comprehensive data on migration

The growth rates in agricultural production and income is noted as low, unstable and disparate across regions. It is evident that there is a decline in the participation of agri-culture production, due to a lack of sector diversification that has led to a shift of labour from the rural to urban, formal or informal sectors. Apart from the above mentioned reason low possession of land combined a lack of irrigation facilities and agriculture failure has prominently led rural people to migration-out of their source area. At the same time that there is rapid development in the modern sector, to absorb surplus labour from the rural area and employ people either in the formal or informal sector.

In addition to the above literature, a recent study by Sharma and Bhaduri (2007) suggest that India might very well be at the "tipping point" in the transition of its agricul-turally dependent population. A large proportion of youth in

17

the countryside are on their way out of agriculture. The above pointed factors have caused rising disenchantment, particularly with-in youth in the village side. The youth pulled themselves out of agriculture because of booming opportunities in other sectors.

From the above analysis it may be concluded that agri-culture is the principal occupation for a majority of the pop-ulation in rural areas. The spread of secondary and tertiary activities is inadequate. Agriculture, in most cases, is in the nature of subsistence farming. Furthermore, average yield per hectare is low, and also subjected to floods and droughts. The over dependent population of the districts fail to sustain themselves solely in agriculture, hence, they prefer to migrate to places outside the districts for employ-ment and livelihood.

The role of Industrialization and Urbanisation in migration:

On the other side, rapid urbanisation and industrialisation have generated more employment opportunities and have created better infrastructure. People migrate to these re-gions perceiving them as greener pastures.

Migration and urbanization are direct manifestations of the process of economic development, particularly in the contemporary phase of globalization. The changing land-scape and growth of economic activity in cities have not only attracted substantial amounts of capital but also a large number of migrants (Das 2008). Hence, most of the non-forced migration may be characterized as pull demand-driven.

Chart 2: Period wise development of Agriculture to Non-agriculture (per annum)

18

Source: Handbook statistics on Indian Economy, Reserve Bank of India.

From the above chart it is evident that because of indus-trialisation, industrial output as a percentage of GDP showed an impressive growth of 5-6 per cent in eighties to 10 per cent in nineties. The increase was more than the percentage share of agricultural output during the same period. Due to increased industrial production, the demand for workers also increased. The supply of labour available in urban areas however, lags behind the demand for it. As a result, additional employment of labour is considered pos-sible only by migration of labour from rural to urban areas (Bhalla and Hazell 2003).

Migrants are attracted by better access to public ser-vices such as electricity, clinics and schools. The ‘bright lights’ of the cities also may be a pulling factor (Sahu and Das 2008). However, although some migrants move for these reasons, most of them respond primarily to economic incentives. People move from poorer areas to wealthier areas for economic gain. Differences in average income or wage levels between rural and urban areas significantly af-fect migration between the two locations.

If economic factors play a major role in determining rural-urban migration, then urbanisation and city growth are clearly determined by those same factors. It follows that urbanisation and city growth cannot be analysed without giving explicit attention to the interaction between rural and urban labour market.

The rural poor migrate to cities even though if they are unlikely to find jobs, but expect to obtain high-wage sector employment. The incentive to wait is the difference between urban and rural wages. Migrants are not only at-

19

tracted by the increased wages due to high-wage employ-ment, they are also willing to be unemployed or accept low wages in the urban labour market for a period of time in the expectation of achieving a high lifetime income.

In conclusion, industrialization and urbanization have resulted in high rates of migration (permanent and tempor-ary). This has occurred despite the fact that many migrants live in appalling conditions and work in the informal sector, which offers uncertain and underpaid work. This is because urban labour markets offer unmatched opportunities to switch jobs rapidly, diversify incomes, and become up-wardly mobile with a very low asset-base and skills. Hence, urban and rural areas may not be considered as two separ-ate isolated units. It is because of this linkage that rural people often migrate to urban areas to acquire better stand-ard of living and employment.

The role of excessive debt burdens:Rural people do not have adequate means at their disposal. Since income from agriculture is too low and in order to meet the scarcity conditions they often have to borrow from informal agencies/money lenders. These agencies supply credit at exorbitant rates of interest which the farmers later find difficult to repay. In order to repay the current loan, the rural people prefer to migrate to places outside the dis-tricts (Deshingkar and Akter 2009).

With this background it is worth mentioning Orissa of a district where migration has taken place because of excess-ive debt. In some districts of Orissa Dalals (local contractor and money lenders) are the main causes of household mi-gration. The Dalals in this region provide advance money which varies from Rs 3000/- to Rs 10000/- depending on the number of working family members, the Dalal lends money to the households. The households cannot get such a big amount of money at a time from any other sources. The in-formal money lending system is too heavy for repayment and there is no provision for the villagers to avail loan from any formal sector for personal consumption needs. In this situation, the Dalals of the brick kiln factories give them chance to take advance money before work. This appears as a great opportunity for repaying the loan taken from the money lenders, releasing the mortgaged land, purchasing bullocks, and to sustain their families. The households then migrate to repay the advance (Deshingkar and Akter 2009).

20

In a village in southern Madhya Pradesh (MP) studied by (Llewellyn 2006), migration to brick kilns has reduced borrowing from moneylenders, and bonded labour. Migrant work appeals to villagers because it presents the chance to earn more money or a larger in-kind payment than they could earn in the village.

Llewellyn states “In a shock situation, people who previ-ously would have had no choice but to submit to a year’s bonded labour for Rs.5000 can now weigh the option of so-liciting an advance from a contractor and migrating in-stead”

The relationship between debt and migration, however, is not straightforward. While some analysts have concluded that migration increases debt levels because of higher ex-penditures during transit and at the destination, others have argued that migration improves the creditworthiness of households and, as a result, allows them to borrow more (Ghate, 2006).

Apart from excessive debt burden other reasons like low production, scarcity of work, irregular work, and low and ir-regular payment have led to movement. However, there are several other determinants that encourage people to mi-grate from rural-urban areas such as expected income, pro-spects of employment, education, and distance from place of origin, networks and costs of migration.

In conclusion, migration may not only be caused by the level of urbanization and the rate of urban expansion or by the ‘pull’ of economic prosperity and opportunity in the cit-ies. Migration is sometimes caused by the push from the rural areas due to low yield, significant changes in the mode of production in agriculture, and land-labour ratio all of which compel migrants to seek a living outside agricul-ture.

3.1.2 Empirical Analysis Drawing on the framework outlined above and based on primary data this section identifies the main factors that drive migration, develops a profile of these migrants and provides an assessment of migrant living and working con-ditions. The information and analysis is organised into five broad categories.

First, I look at factors that cause migration. Then, I move on to develop a migrant profile by using variables such as age, sex, caste, religion, district, education level, marital status, years of migration, and family and child mi-

21

gration status. To assess the living conditions, I use vari-ables such as migrant annual room rent and room share in comparison to occupation, income level and work hours which reflect their living arrangement at destination place. To assess work conditions, I use variables such as present occupation, income, and work hours, and in addition, work place conditions. Finally, I examine migrant-illness to work and living conditions.

Hence, all the above mentioned details provide an in-sight into the context of internal migration from the per-spective of remittance sending individuals and their socio-economic conditions and how this context would possibly affect their future working ability.

In the following section I use univariate, and bivariate frequency analysis followed by multivariate regression to specifically examine the factors that influence illness. While using multivariate regression I begin with a model specific-ation for migrant illness and follow that with an empirical analysis.

Model specification:In the present model I use V, a dichotomous variable, to de-note the presence/absence of illness among the Oriya mi-grant. V may be treated as a function of variables capturing the overall socio-economic position of a migrant (XM) in the destination place. The other explanatory variables in this model are the age of the migrant (XA), the city of migration (XC), the present occupation of the migrant as a machine operator and working in ports in Surat and Gandhidham re-spectively (XTFO, XWP), and a vector of an additional explanat-ory variable (XO). Thus, illness may be represented as,

V= XC βC + XA βA + XTFO βTFO + XWP βWP + XO βO + є ………….. [1]

The βs are the coefficients to be estimated and є repres-ents unobservable factors which may influence illness. Based on the assumption that є follows a normal distribu-tion this equation may be estimated using a probit model.

In operational terms, V captures the incidence of a mi-grant falling sick. The overall socio-economic condition of the migrant is captured by the city of migration, the age of the migrant, the present occupation of the migrant espe-cially in Surat and Gandhidham; the education level of the migrant, years of migration, work and living conditions (which include variables like availability of water, sanita-

22

tion, and places to eat, work hours for each of the above mentioned occupation, chemicals exposed and instruments used at the work place); and, lastly, the wages of the mi-grant.

Variable SpecificationIn this section I describe the variables and also explain the reason for specifying these variables in the model. The fol-lowing independent variables are included in the model for analysis: City (Surat and Gandhidham) – The cities of Surat and Gandhidham are the labour receiving states. This is a dummy variable that represents the following: the migrant living in Surat is 1, and 0 otherwise, and the same is applic-able for Gandhidham. Using this variable, I will try to ana-lyze whether choice of city is an important factor for an in-crease or decrease in migrant illness.Age – The average age of a migrant is around 27 years. The migrant usually migrates at a very young age but after a certain period that is in the long run the migrants above a certain age tend to return back to their villages. Hence, in this aspect I want to look at whether age has any significant relationship to migrant illness.Present occupation – The occupation of the migrant is a dummy variable. For analysis, I consider the machine oper-ator and working in port. The other variable is termed as a reference category. The migrant mainly works as a machine operator, or in port or other services. I want to look at whether the choice of occupation has any influence on a mi-grant being ill or not.Education level – Education of a migrant is also a dummy variable. I have created five categories. They begin with no education level, followed by primary, secondary, middle, high and other education levels. Here I want to look at whether education level has any significant relationship to a migrant falling sick or not. Years of migration – It has been observed that the aver-age years of migration have been around 7 years. Hence, it has been included in the model to show whether it has any influence on migrant illness or not.Living and working conditions – In this category, I in-clude variables such as water, sanitation and places to eat to look at whether variables have any significant influence on migrant illness. Form the descriptive analysis it is evid-

23

ent that most of the migrants do not have access to proper amenities either at their work or living place. Therefore, I have included them in my model to show a significant rela-tionship to migrant illness.Chemicals and instrument – This variable is a dummy variable. The instruments used are Panna Paanch and Sat-tal. I look at whether the usage of these instruments makes a migrant fall sick or not. From the analysis it is observed that these machines produce a lot of noise and heat. As a result, I have included these variables in my model.Work hours - It is observed that on an average a migrant works for 11 hours a day under strenuous conditions and gets only one unpaid leave a week. If it is a peak season, mi-grants work for even 16 to 18 hours a day. Hence, work hours have been included to show whether they have any significant relationship to migrant illnesses.

In my empirical work I explore the sensitivity of my results to other specifications like Ordinary Least Square (OLS) to ensure robustness in my results.

Rational choice for Probit Model:In this paper, I explain the dependant variable illness which has a binary outcome: whether a migrant falls sick or not. It takes only two discrete values: 0 and 1. This Limited De-pendent Variable Model (binary dependent variable model is one of them) can be solved in different ways. The Linear probability model (LPM) and the Logit and Probit Models are the possible alternative ways that can be used to solve binary dependent variable model.

The LPM model is very easy to estimate and explain but it has some limitations that make it less useful and unpopu-lar in estimating a limited dependent variable model. First, for example, sometimes this model provides a predicted value that is greater than 1 or less than 0. Second, the par-tial effect of any explanatory variable is constant and the outcome of the model is a probability that cannot be lin-early related to the independent variables for all their pos-sible values. Third, the variance is not homoskedastic in a linear probability model. The problems of estimating binary choice models can be solved by using the Logit or Probit model. Conceptually, both are the same. The only difference is in the distribution functions: the Probit model choose the standard normal distribution and Logit chooses the logistic distribution. Sometimes econometricians like the Probit

24

model due to its standard normal distribution, although it is quite difficult to estimate and explain.

The Probit model can be derived from a latent variable model, such as

iii xy *

[2]

) ( 0' 0

) ( 0' 1*

*

valuesunobservedthexyif

valuesobservedthexyify

iii

iiii

Where and are assumed independent and and *

iy denote a latent dependent variable that is unob-served and the observed value of amount of remittance received by the ith households.

Equation (1) tells us that is either one or zero. We point here also that denotes the vector of remittance de-termining variables for the ith households and the error term.

Now,

Pr{yi = 1| x}= Pr{yi >0} = Pr{ }= Pr{ > -}= 1 - =

The maximum likelihood function is then

ii yii

yii

N

i

xyxyxyL

1

1

};0Pr{};1Pr{),(

Using the maximum likelihood method we can find the Probit estimate of marginal effects of changes in independ-ent variable which is expressed as:

kiik

i xxx

3.1.3 Empirical Results

a. Factors driving migration and migrant profileIn this section, I provide analyses on the motivation for mi-gration and general information on the migrant’s profiles

25

from the perspective of the remittance sending individual. The analysis in this chapter is on the destination place, Surat and Gandhidham.

Table 4: Motivation for Migration: Extended Household

S.No Push Factors % when compared to total No of HH

1 Unemployment 97.22 (280)2 Education 53.47 (154)3 Poor Industrial

Growth53.13 (153)

4 Repayment of Debt 51.39 (148)5 Marriage 31.25 (90)6 Agricultural Failure 19.79 (57)7 Natural Disaster 1.04 (3)8 Political Instability 1.05 (3)

Pull Factors9 Better Wage Rate 33.10 (95)10 Better Opportunities 10.80 (31)

Source: Fieldwork, 2010 Notes: Figures in parentheses are number of respondents indicating frequency of the listed factors that motivated them to migrate.

Table 4 divides the various reasons for migration into push and pull factors that encourage migration from the source areas to the destination points. The figures show that the total sample of 288 individuals currently living in Surat and Gandhidham provided 10 reasons for migration. The major reason for migration from the source area was unemployment 97.22% or 280 of 288, individuals in my sample migrated because of unemployment. The next reason for migration from the source area was better edu-cation to their children about 53.47% or 154 individuals have migrated to provide better education to their children, while 53.13%, or 153 individuals cited poor industrial growth as a reason for migration. Repayment of debt was the reason for migration from the source area 51.39% of the total individuals in my sample. In the destination place, people were attracted to the location due to a better wage rate (33.10% of the sample) and employment opportunities (10.80% of the sample). Migrants usually have more than one reason to migrate. Migration does not occur only be-cause of unemployment or repayment of debt, but rather because of confluence of all these factors.

The figures for causes of migration is in accordance with other studies. For example, one study (Sahu and Das 2008) reveals that the lack of employment opportunity accounted for 21% of migration, while 15% reported to have migrated due to repayment of debt. Other factors such as family in-come, poor financial condition and the land-man ratio (acres of land per person) aslo led to out-migration. In the

26

case of pull factors, better employment opportunities ac-counted for 53% of migration, while higher income was about 6%. However, there is a sharp difference in the data on certain push and pull factors due to the difference in the sample size. Both the studies analyse on similar factors and conclude that migration from the source to the destination place is caused by a confluence of push and pull factors.

Table 5: Age distribution of migrants

Age in years Frequency Percent16-20 44 15.2821-25 96 33.3326-30 56 19.4431-35 49 17.0136-40 27 9.3841-47 12 4.1748-54 4 1.39

Source: Fieldwork, 2010

Table 5 provides information on the age distribution of migrants. The figure shows that out of 288 individuals living in Surat and Gandhidham, 15.28% belong to the age group of 16-20 years, which includes child migrants. The highest and lowest number of migrants was found between the age group of 21-25 and 48-54 years, respectively. The youngest and eldest age of migration was around 16 years and 54 years respectively. The median age of these migrants was around 27 years. As the age group increases beyond 31-35 years there is a fall in the number of migrants staying in the destination place. Two reasons were mentioned for out-mi-gration: first, accumulation of savings and, second, age. After the age of 40 years most of these migrants are unable to work for long hours (12 hours). As a result they return to their villages with or without savings. The younger mi-grants, therefore, replace the elder migrants. After a partic-ular period of time when the size of remittance reduces and expenditures increase, the workers decide to pull them out of the urban labour market. The figures on the age of re-spondents are in accordance with other studies. For ex-ample, a study by Sahu and Das 2008 reveals that the youngest and oldest age of migrants was 18 and 51 years, respectively, and the median age of these migrant workers were around 30 years. The study also suggests that elderly migrants often find it difficult to work at the pace and speed that the younger migrants may adopt while working. Fur-thermore, increasing age is likely to reduce income and size of remittance. However, there is a difference in the data on the index of the minimum age of migration, which relates to child migration.

27

Table 6: Distribution of Sex, Caste and Religion of migrant:

Source: fieldwork, 2010

Table 6 presents general information on the distribution of sex, caste and religion in the event of migration. Out of the 288 migrants, invariably all of them are male migrants and, in the case of religion, the only group that has mi-grated out are the Hindus which have a frequency of about 288. From the table it is important to note that the highest migration is generally from Other Backward Caste, which account for 74.63%, followed by the General caste, which accounts for 24.31%. The Schedule Caste accounts for 0.69% and the remaining 0.35% belongs to the Schedule Tribe. The lowest levels of migration are observed in Gen-eral caste and Schedule caste and tribe. People belonging to higher castes often find it difficult to opt for lower levels of jobs, which tend to denigrate their social status at their native locations. However, they are able to conceal their job status at their native ends even while working as a wage la-bourer which does not match their social positions. There-fore, we find lower indices of out migrants in the General caste. In the case of the Schedule caste and tribe, lower levels of aspirations and the availability of ‘low paid un-skilled jobs’ in the native itself may perhaps be attributed to the lower rate of migration from SC and ST community. A critical point, however, could also be the extent and intens-ity of poverty, which is pertinent among SC and ST which reduces the mobility when compared to other caste. The vulnerable poor are often unable to migrate out to farther locations, often due to their inability to afford the costs as-sociated with migration and the lack of community support, especially in the destination point.

However, other studies conducted by the Department for International Development (DFID), which funded the

Migrant SexSex Frequency % when compared to no. of house-

holdsMale 288 100.00Migrant Caste GC 70 24.31SC - -ST 2 0.69OBC 215 74.63Others 1 0.35Migrant ReligionHindu 288 100.00Muslims - -Christians - -Others - -

28

Western Orissa Rural Livelihood Project (WORLP) in the districts of Nuapada and Bolangir show a dominance of STs and SCs in migration streams (Panda 2005). Along with these communities, backward castes are also heavily rep-resented in migration. The WORLP research also shows that although migration among females is lower than males, a significant number of females migrate (39%) in the Khariar block of Orissa. This study also observed that child migration constituted 14.89% of the total number of migrat-ing individuals in the sample. There has also been a shift in the caste composition of migrants: earlier studies (e.g. Nayak 1993, Sahoo 1993) note that a majority of migrants belonged to the other castes (i.e. upper castes) and a much smaller proportion belonged to the Scheduled castes and tribes. Furthermore these studies noted that most migrants belonged to the small and marginal farmer category. How-ever, toward the end of the last decade, migration streams had a high proportion of SCs (Samal 1998).

However, there are sharp differences between present study and other especially with respect to the caste and sex of the migrants. In my data, the representation of the SC and ST are low or nil but in the case of WORLP, the findings are different. This suggests that the distribution of the SC and ST in these districts is very uneven. The same is the case with the sex of migrant, depending on the location of the study and the nature of job, one could find females as migrant workers.

Table 7: Distribution of State, District and City of migrant:

Migrant State

State Frequency % when compared to no. of HHsGujarat 288 100.00Destination DistrictSurat 156 54.17Gandhidham 132 45.83Destination CityLarge 156 54.17Small 132 45.83

Table 7 presents general information on the preference and distribution of migrants according to the state, district and city. The table has been broadly categorized into 3 sec-tions:-

1. Distribution of migrants according to state2. Distribution of migrants according to choice of their

destination districts 3. Distribution of migrants according to cities.

29

Out of the total sample size, 288 respondents were loc-ated in Gujarat (frequency 288), 156 were located in the in-dustrial city of Surat and the remaining 132 were located in a small city of Gandhidham and the port of Kandla. The dis-tribution of migrants in districts and city is the same. In Surat, migrant workers are more likely to get a job either in the powerloom or diamond industries. Surat provides an ex-tensive job market for the migrant, even in the case of re-trenchment, because of the huge presence of powerloom in the city. In Gandhidham, migrants get a job either in the garment industry or at the port of Kandla. In the case of re-trenchment, finding a new job in Gandhidham is difficult as the workers have to wait until the port supervisor calls them back to work. Moreover, the movement of migrant workers at the destination depends on the presence of their fellow community or state people. Lastly, these migrant workers have no fixed place of work because the Surat job market provides work for 9 months and for the remaining 3 months, they either go back to their villages or Mumbai or another place for work. This results in circulatory move-ments.

Table 8: Education level of migrant

Education level Frequency PercentPrimary 73 25.35Middle 90 31.25Secondary 92 31.94High 14 4.86Other 4 1.39No Education 15 5.21

288 1.0000

Source: fieldwork, 2010

Table 8 presents information on the Oriya migrants’ educational level. About 31.94 % of the migrants’ have a secondary level education, 31.25% of the migrants’ have middle level of schooling, 25.35% of the migrants’ have primary level of education and a small have high and other level of education and some have no education. In conclu-sion, the general trend is that people who a have low level of education tend to be more mobile and end up getting jobs in the informal sectors. The same may not be applic-able, however, to those who have relatively better educa-tion. If the migrant has a low level of education, then she/he stands no opportunity of getting a job in the formal sector and hence resorts to informal sector for meagre susten-ance. According to a study by Srivastava (2003) suggests

30

that bulk of migrant work force in India has little or no edu-cation. The findings of this data suggest that the majority of the migrants lack minimum education and hence, end up getting jobs in informal sector.

Table 9: Details on Marital Status, Family Migration and Child Migration

Migrant Marital Status

Marital Status Frequency % when compared to no. of HHMarried 116 40.38Unmarried 172 59.72Migrant Family StatusOrigin (Orissa) 271 94.10Destination (Gujarat) 17 5.90Child MigrationYes (Children) 99 34.38No (Adult) 189 65.63

Source: fieldwork, 2010

Table 9 presents information on marital status, family migration and child migration17 at the destination place. First, in case of marital status, an unmarried migrant is more mobile and flexible than a married migrant. The move-ment of migration among married migrants is restricted to 40.38 %, but in case of unmarried migrant the movement is about 59.72 %. This is a higher scale in comparison to the married migrants. Second, only 5.90 % migrated with their family. The remaining 94.10 % live with their own com-munity people on a shared basis. It is easier for the unmar-ried people to migrate because of their marital status, the types of jobs they obtain, and the different nature of jobs, living arrangements and costs involved.

In summary, migration from rural to urban areas is not only stimulated by push factors but is also triggered by pull factors. It is evident from the inquiry that actual urban-rural wage differentials and employment opportunities emerge as important pull factors for out-migration. The analysis then moves on to assess migrant profiles. In the case of age, over a period of time, worker productivity tends to decline and these workers are not absorbed into the urban labour market. They are usually replaced by a younger generation of workers migrating from various parts of the country. Hence, it is clear that these migrants return home and survive on subsistence farming or enter into non-farming activity at the source place. With respect to sex, caste and religion, it is prominent that the only

31

group that has out-migrated is Hindus and invariably all are male migrants. Among the Hindus, only Backward caste people have migrated. Movement among the other castes is restricted. This suggests that migration largely depends on networks at the destination place. The analysis then moves on to highlight the migrant choices for a particular destina-tion place, which depend on work availability, employment opportunities, wages and networks at the destination place. The analysis on education level provides a clear under-standing as to why migrants get absorbed into the urban in-formal sector.

17 A child means every human being below the age of 18 years unless under the law applicable to the child, majority is attained earlier. In this context since are particularly looking at child la-bour it is appropriate to use the Prohibition and Regulation act, 1986 that define a child as a labour if she/he is employed below 14 years of age.

The majority of these migrants are at a 10th or 12th

pass/fail education level. Very few of them have high quali-fications such as graduation, post graduation and diplomas. Since the majority of these migrants have low educational qualifications, they are readily absorbed into the urban in-formal labour market. Finally, in the case of child migra-tion, the proportion is larger when compared to the total sample size. The data analysis also reveals that child migra-tion is ascending (34.38%) because of rising poverty, family breakdown, indebtness, mother- led households and a lack of income opportunities in the source areas.

b. Living and working conditions of migrantsIn the above section, I briefly discussed what motivates mi-gration and then moved on to develop a profile of the mi-grant. This section of the analysis examines the living and working conditions of migrants. To assess living conditions I use variables such as present occupation, working hours and income levels which determine the living arrangements among the Oriya workers in Surat and Gandhidham. Tables 7-9 are univariate and Tables 10 & 11 are bivariate tables.

Table 10: Migrant occupation in informal sector at Surat and Gandhidham

Present occupation Frequency Percent

Machine operator 154 53.47Working in port 105 36.46

32

Saree work 11 3.82Spinning mill 6 2.08Bobbin Operator 3 1.04IFFCO worker 2 0.69Welder 2 0.69Sawmill operator 1 0.35Supervisor in IFFCO 1 0.35Tailor 1 0.35Carpenter 1 0.35Cycle shop 1 0.35

288 100.00

Source: fieldwork, 2010

Table 10 provides information on the present occupation of migrants. The major source of occupation in Surat is the powerloom and diamond industry. In Gandhidham, mi-grants are mainly engaged in jobs associated with knitting or at the Port of Kandla they work as loaders and unload-ers. Out of 288 individuals, 53.47% work as machine oper-ators and 36.46% work in the port of Kandla. The remaining individuals work as carpenters and welders. In conclusion, among 288 respondents, only one migrant was an entre-preneur and the rest were daily wage labourers. The overall figures on the occupations of Oriya workers appear to be in accordance with other studies. For example, a study by (Sahu and Das 2008) reveals that Oriya migrants in Surat account for 79.5 % of the work in Textile related jobs. The data confront on the same issue. However, the variations in the data are due to the size of the sample and the fact that the present study was conducted both in Surat and Gandhidham whereas the study conducted by Sahu and Das was concentrated only on Surat.

Table 11: Migrant working hours

Working hour Frequency Percent8-12 271 94.1014-16 15 5.21>17 2 0.69

Source: field work, 2010

Table 11 provides information on working hours. If a mi-grant is employed as a machine or bobbin operator in the powerloom industry, he works for 12 hours. In Gandhid-ham, on the other hand, a worker works for 8 hours, which varies depending on the quantity of material at port. Among 288 individuals, few migrants work for more than 14-18 hours or even 24 hours a day. In such instance it is was dis-covered that for 15 days they work for such prolonged hours and for the remaining 15 days they rest. Out of 288

33

individuals, 15 migrants were found to be working for such prolonged hours while the remaining worked for 12 hours. The findings of my analysis concur with other studies. For example, a research done by an NGO (Sansristi, 2007) de-termine that families that migrated from Orissa to Andhra Pradesh were engaged in brick kiln work18. They work for 12 to 13 hours in a day trying to make as many bricks as possible. They will also work for 16 hours a day to complete the target. In the present study on an average a migrant works for 11 hours a day. Another study by (Deshingkar and Akter 2009), suggests that a worker in the powerloom sector in Surat works 12 hours in a day with a 30 minute break. If the other workers are absent, then co-workers may have to work for 24 hours without a break. In addition, they are given only one unpaid holiday a week. Hence, on an average, irrespective of the place of migration, the mi-grant continue to work for prolonged hours under exploitat-ive conditions.

Table 12: Wages before and after migration (last one year) per annum

Wages before migration Frequency PercentNo Wages 155 53.821200-2100 78 27.082101-4500 29 10.074501-10000 11 3.8210001-38000 15 5.21

Source: fieldwork, 2010

18 In this work the whole family (minimum three members) is en-gaged as a team. If the family has only two members, then the family takes one of the other relatives to complete the team. The work in which they are engaged is called as chhanchua or pa-thuria (brick making). In brick making there is a division of la-bour between males, females and children. The males are en-gaged in digging the clay at the work place, mixing the clay and making the bricks with the help of frames. The women are in-volved in making the clay dough. Both women and children shift the raw bricks to dry under the sun and then keep them in the raw brick bhati.

Wages after migration Frequency Percent57600-64800 66 22.9245620-51000 52 18.0665700-77000 50 17.3679200-96000 34 11.8137500-45000 33 11.4654000-57543 29 10.0724000-36000 19 6.60>96550 5 1.74

Source: Field work, 2010

34

Data on income is based on the last one year. Out of 288 individuals155 (or 53.82%) had no wages before migration because they were working in their own field or unem-ployed. 78 individuals (or 27.08%) individuals had an an-nual average income of Rs 1200/- to Rs 2100/-. Only a few individuals (5.82%) have an annual income of Rs 10000/- to Rs 38000/-. After migration, however, the same set of indi-viduals experience an increased income when compared to the income earned at the source place. Around 22.92% of the migrants have an annual income of Rs 57600/- to Rs 64800/- while, 18.06% of the migrants have an annual in-come of Rs 45620/- to Rs 51000/- and 17.36% of the mi-grants have an annual income of Rs 65700/- to Rs 77000/-. In most of the cases, it is clear that migrants do experience an increased and constant flow of income which they remit back to their families. Apart from increased income, those migrants who were not employed also got employment op-portunities in the destination place. This confirms the the-ory that higher wages and better employment opportunities are pull factors for migrants in the destination place.

In conclusion, on average, migrants earn per day earn around Rs165/-, Rs 5000/- per month and Rs. 50000/- per year. While calculating income only ten months were taken into account because, the powerloom in this region would be closed for two months of the year. Therefore, only 10 months were used to calculate the migrants’ income. The same is applicable even in case of Gandhidham. In order to substantiate the above findings, I also look at other studies done by Sansristri, 2009, in which wage payment for an adult worker gets Rs 100/- a week. A child gets Rs 70/- a week. Depending on the number of workers in a family, the family could get around 270/- to 370/- per week. The present study found no difference in the wage level between adult and child migrants. All of them earned between Rs 165/- to Rs 267/- per day and the wage payment was done on weekly basis.

Table 13: Living Arrangement of migrants according to city and room rent (last one year) per annum

Rent share per migrant District at destination place

Surat Gandhidham

400-1,620 (18) 11.53 (2) 1.511,800-2,880 (57)36.53 (5) 3.783,000-4,000 (56) 35.89 (58) 43.934200-6168 (6) 3.84 (51) 38.636600-12000 (11) 7.05 (16) 12.1213,200-24,000 (8) 5.12 (0) -Source: fieldwork, 2010

35

Table 13 provides data on the migrants’ living arrange-ments at the destination place and per annum expenditures in room rent. The information on room rent is based on the last one year. Out of 288 individuals, majority of the mi-grants’ room rent is between Rs 1800-2880 and their fre-quency is about 57, followed by Rs 3000-4000. Their estim-ated frequency is about 56 (35.89%). The rest of the mi-grants’ room rent share ranges as follows: Rs 400-1620 (18), Rs 4200-6168 (6), Rs 6600-12000 (11) and Rs 13200-24000 (8) respectively. In a small city like Gandhidham mi-grant room rents range from Rs 3000-4000 and their fre-quency is about 58, followed by 4200-6168 frequency which is about 51. Following figure gives us information about rest of migrants living arrangement Rs 400-1620 (2), Rs 1800-2880 (5), Rs 6600-12000 (16) and Rs 13200-24000 (0). When compared to Gandhidham, the per annum room rent in Surat is comparatively lower.

Table 14: Living arrangement of migrant according to city and room sharing basis

Room sharing Surat Gandhidham

0-3 (23) 14.83 (12) 9.024-7 (86) 55.48 (52) 39.098-12 (46) 29.67 (64) 48.1217-20 (0) 0 (5) 3.75Source: fieldwork, 2010

Table 14 details the living arrangements of the migrants according to the number of people sharing the room. In a large city like Surat the frequency of 4-7 migrants living in a room is about 86 (55.48%). The frequency of 8-12 mi-grants sharing a living is about 46 (29.67%). The remaining migrants are in the range of 0-3 per room (23) and 17-20 per room (0). In Gandhidham, the frequency of 8-12 mi-grants living in a room is about 64 (48.12%), followed by 4-7 living in a room is about 52 (39.09%). The remaining indi-viduals are in the range of 0-3 people to a room (12) and 17-20 people to a room (5). In comparison to Surat, groups liv-ing together in Gandhidham which has relatively are charged relatively higher rents by the landlords. The higher rent forces a migrant to be very economical and live in dingy places which affect their health, mentally and physic-ally. In addition, landlords treat migrants as an income source and really do not work to provide better living condi-tions to them.

To substantiate the findings and information provided in the present study on room rent shares and living arrange-ments, I refer to other studies done by Sahu and Das, 2008, which show that the living arrangements of these Oriya workers are interlinked to with extent of joblessness. From

36

the study it has been determined that Oriya worker on aver-age spend around Rs 187 to Rs 217 per month on rent. The findings of my study suggest that on average spend around Rs 142/- per month on rent. This sharp difference is due to selection of the place of study. From the analysis it is evid-ent that rent in Gandhidham is comparatively than rent in Surat. Hence, this explains the variation in the rent between the two studies and also confirms that, to minimise expenses, migrants tend to live in large groups of 10 to 12 people, according to Sahu and Das, 2008, and in the present study migrant live in groups of 7 to 10 people.

In summary, the information provided in this section forms a platform for assessing the living arrangements of Oriya migrants at the destination place. From the above tables it is clear that on average a migrant earns around Rs 59431/- per annum and that the rent paid per annum is about Rs 4260/-, which is around 7.17% of their total earn-ings. When compared to their current earnings at the des-tination place, the percentage of amount spent on house-rent is comparatively high. Hence, to minimize their ex-penditure they tend to live in large groups of 7 to 10 people. The migrants belong to the same occupation but have different work shifts. They stay together so that they can accommodate a large number of individuals in a small room. However, in the long-run such living arrangements tend to have a negative effect on their mental, physical, physiological and emotional balance.

c. Migrant-illness to living and work conditionsAt the beginning of the preceding section information on in-come, work hours and present occupation was analysed. Thus, in this section I will move on to analyse correlations between work hours, income and migrant-illness19. Initially I provide univariate and bivariate tables and then examine the multivariate regression.

This section describes migrant-illness in three broad categories: work hours, living condition and availability of basic facilities at the work place. Apart from work hours and income other variables include sanitation, drinking wa-ter facilities and a place to eat at both places locations.

Table 15: Availability of facilities at work place

Working place Facility Frequency PercentDrinking water source 121 42.01Sanitation facility 108 37.50Facility to eat 99 24.38

Source: fieldwork, 2010

37

From the above table it is clear that most of the mi-grants working in the urban informal sector do not have ac-cess to basic facilities at their work place. Out of, 288 indi-viduals, 57.99% do not have access to drinking water at their workplace. In the case of sanitation facilities 62.50% are not provided sanitation facility at their work place and 65.63% have no proper place to eat at their work place.

Table16: Availability of facilities at Living place

Water Source Frequency PercentTap Water 241 83.97Water board vehicle 4 1.39Common drinking water 42 14.63Sanitation Facility Frequency PercentYes 233 80.90No 55 19.10

Source: fieldwork, 2010

19 In this study illness is accounted on the basis of migrant falling sick in the last one year. Illnesses mean disease of body or mind, poor health or sickness. It also includes aspects of physical, emo-tional or physiological condition and function of a person that are diminished or impaired compared with that person previous con-dition.

The above table provides information on drinking wa-ter sources and sanitation facilities available at the destina-tion place. From the above table it is observed that out of 288 extended households, 83.97% have access to tap drink-ing water, 14.63% have access to common drinking water and 1.39% have access to water board vehicle drinking wa-ter. In case of sanitation facility, 80.90% of them have ac-cess to sanitation facility but the remaining do not have ac-cess to sanitation facility.

Table 17: Health Problems faced by migrants (last one year)

Health Problems Frequency PercentCold and Fever 194 63.67Malaria 72 25.00Typhoid 9 3.13Jaundice 7 2.43Eye Problem 2 0.69Others 4 1.39Total 288 100.00

Source: fieldwork, 2010

The above table provides information on migrant health problems which arise because of the living and working conditions. In this case, for the purpose of analysis, I have taken only drinking water and sanitation facility in both en-vironment and eating facility at the work place. From the

38

above table it is observed that 63.67% suffered from cold and fever, followed by malaria which is 25.00% and the oth-ers suffered from Typhoid, Jaundice and eye problem.

To substantiate the information provided on migrant-ill-ness I also consider another study done by Deshingkar and Akter 2009. It informs on social and basic facilities (hous-ing, food and health status) available for Oriya workers at brick Kilns at Andhra Pradesh. The workers live in tempor-ary low-roofed and thatched houses in the worksite. The water provided is not suitable for drinking purposes and there is an absence of toilet facilities at the work place, as well as in the living place. The study further emphases that migrant households purchase low grade rice and vegetables for consumption. They often suffer from dysentery and other nutrition deficient diseases. The water discharged, by the factories making bricks, makes the migrant suffer from various skin disease and colic disease. A confluence of all of these factors affects the health status of the migrant. The findings of my study are concurrent with the study done by Deshingkar and Akter 2009. The variation of, and drawback to the present study is the exclusion of food habits that could also possibly affect the migrant-illness. Both the stud-ies suggest that work place and living conditions have an impact on the types of illnesses faced by the migrant.

In summary, the finding of my analysis highlights mi-grant-illness in relation to the work place as well as work-ing and living environments. The previous analysis on mi-grant-illness was in relation to work hours, whereas in this case I look into the living and work place conditions that further impact a migrant becoming ill or not. From the ana-lysis it is evident that workers fall ill not only due to work hours but it is also combination of other factors available fa-cilities at work place, living environment and also living ar-rangement. Most of the illnesses are cold and fever but, wa-ter borne disease like (Jaundice, Malaria and Typhoid) are also prominent. Illnesses can be due to unsafe working con-ditions in the powerloom sector. In the case of Surat the worker are exposed to heat and sound produced by ma-chines, as well as, long work hours which makes them fa-tigue and less productive over a period of time. As this pro-cess is continued several times over and over again, they become chronic patients which become severe after coming back to their native villages. Therefore, these factors do have a negative impact of migrant future working ability.

Table 18: Multivariate Regression on migrant illness

39

VARIABLES OLS d-Probit*Surat -0.223*

(0.116)-0.295**(0.126)

Age 0.00338 (0.00426)

0.00135(0.00319)

*Present occupation (TFO- Ma-chine operator)

-0.234**(0.105)

-0.0932(0.0652)

*Present occupation (Working in Port)

-0.117(0.125)

-0.130(0.153)

*Middle -0.321***(0.115)

-0.0493(0.0671)

*Secondary -0.257**(0.114)

0.0457(0.0646)

*High -0.327**(0.148)

-0.108(0.139)

*Other -0.386*(0.225)

-0.251(0.256)

Years of migration -0.0154**(0.00605)

-0.0102**(0.00439)

*Living water tap -0.0684(0.0791)

0.0401(0.0798)

*Living water board vehicle -0.0237 (0.213)

0.0566(0.0654)

*Living sanitation (Yes) -0.239*** (0.0721)

-0.266***(0.0943)

*Work sanitation (Yes) 0.00175(0.107)

0.0521(0.0639)

*Work water (Yes) 0.133(0.0912)

(0.0550)

*Work eat (Yes) 0.00589(0.109)

0.00925(0.0815)

Working hours -0.00639(0.0128)

-0.0136(0.0137)

Chemicals exposed 0.166**(0.0806)

0.118**(0.0534)

*Instrument (Sattal) 0.0611(0.152)

0.0750(0.0762)

*Instrument (Heavy goods) 0.0465(0.0831)

0.0405(0.0682)

*Instrument (Panna Paanch) 0.324**(0.159)

0.186***(0.0633)

Constant 1.054***(0.276)

Observations 287 287R-squared 0.317Standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1

*Dummy Variables

At this stage to estimate the working and living condi-tions, of migrant which may lead to an increase in probabil-ity of migrant illness I use the Probit model for binary choice dependent variable. Since the Probit coefficient shows the effect of independent variable on the unobserved latent variable, I focus on its marginal effects.

From the marginal effects of Probit model presented in the above table, it is observed that depending on the choice of destination city like Surat reduces the probability of mi-grants falling sick by 29.50 percentage points. This means that migrants who live in Gandhidham are more likely to fall sick than those migrants who live in Surat and the coef-

40

ficient is significant at a 5% level of confidence. Therefore, the choice of place to migrate and living environment de-termines whether a migrant would fall ill or not.

Years of migration is also an important factor that de-termines the probability of a migrant falling sick. It is ob-servable that minimum years of migration has been a period of seven years, and therefore, over a period of time it is possible that migrants adopt to the work and living con-ditions of Surat and Gandhidham or they may also acquire new skills in the destination place, as that results in the change of occupation as that could reduce impact on a mi-grant falling sick. From the analysis it is observed that as years of migration increases by one year then the probabil-ity of a migrant falling sick reduces by 1.02 percentage points and the coefficient is significant at 5% level of confid-ence.

Living condition especially sanitation is an important factor that determines the probability of a migrant falling sick. In reference to the above table, if the migrant has ac-cess to sanitation facilities then the probability of a migrant falling sick reduces by 26.66 percentage points. Thus, it im-plies that when a migrant does not have access to sanita-tion facilities in the living place then is more likely to fall sick when compared to a migrant who has access to sanita-tion facilities in the living place. In the field it also observed that most of the migrant sanitation facilities are not prop-erly maintained, as that could also cause illness. For ex-ample in a locality like Surat it was observed that a housing unit which consists of 4-6 rooms on each floor with 7 to 10 migrants living in each room. If there were three floors then there were only three toilets and sanitation facilities were available for all the migrant living in that housing unit. In most of the cases the toilets were highly unhygienic: this could also be an influencing factor for a migrant to fall to sick.

If a migrant is exposed to chemicals at his work place then he is also more likely to fall sick when compared to those migrants who do not have chemicals at their work place. From the above table it is observed that if a migrant use chemicals then the probability of falling sick increases by 11.88 percentage points and the coefficient is significant at 5% level of confidence. This was observed only in Surat because; some migrants worked in dying units and were not given any precautions while using the chemicals. The mi-grants usually cook and eat without properly washing their hands. The usage of chemicals also changes the colour of

41

their hands. These factors have adverse affect on their health causing skin disease.

If a migrant uses Panna Paanch 20 at their work place the probability of falling sick increases by 18.60 percentage points and this coefficient is significant at 1% level of con-fidence. The coefficient also suggest that when compared to another instrument like Sattal and heavy goods when a mi-grant uses Panna Paanch, he is more likely to fall ill when compared to other instruments used by the migrants. Panna Paanch is a machine used in the textile industries to weave the cloth, the heat and noise produced by the machine is enormous and the number of machine in a room is around 5-7 or more depending on the space available. Since most of the migrants work for 11 hours on a shift basis they sleep under the machine. The probability of falling sick when working under such conditions is greater.

In conclusion, this section of analysis shows that mi-grant-illness is a confluence of factors, showing that the choice of destination city, occupation of the migrant, living and working condition of migrant and instrument and chemical used at work place increase the likelihood of a mi-grant will fall sick. From the descriptive analysis it is ob-served that on average a migrant work for 11 hours, and it has been observed that they work under inhuman condi-tions, when included in the model the variable turned out to be an insignificant coefficient. This shows that apart from work hours there are other variable which have significant impact on migrant illness. Therefore, by using the probit model establish a relationship was established between mi-grant illness and factors that causes migrant-illness. Mi-grant-illness also has possible influences on the migrants future working ability, as these factors may limit the mi-grant from staying a prolonged period of time. This also res-ults in forced out-migration from the destination place be-cause of low productivity when compared to young mi-grants and reduces the size of the remittances being sent back home.

The major focus of this paper was on migrant living and working conditions as well as on migrant illness. In this sec-tion I looked at the gains of migration from perspective of increased and stable income as well as employment. The table below captures information on income and employ-ment at the source and destination place.

Table 19: Average Per capita income of migrants’ households at Origin by income source

Income and working details

Total working days Per capita income

42

Non Migrant per capita income (Agri + Non-agri)

210 3720.00

Migrant per-capita in-come (Agri + Non agri)

210 3720.00

Migrant per capita in-come from remittance

320 4659.00

Total per capita income for Migrant hh

- 8378.00

Total per capita income for Non-migrant hh

- 3720.00

The above table provides information on per capita in-come from agriculture and non-agriculture activity. It is evident that income from agriculture and non-agriculture is lower when compared to per-capita income from remit-tance. The total per-capita income for a family with remit-tance is around Rs 8378.33/- whereas, if the family has no migrant then the total per-capita of that family is around Rs 3720.00/- this shows that those families that have a migrant have a relatively higher amount of income in comparison to those that did not choose to migrate. It is also evident that the total number of working days in the source areas is lesser when compared to total number of working days in the destination place. This also reflects on the stability on the flow of income with families that have migrants. In con-clusion the major factor that triggers migration is low wages and unemployment. The net gains from migration are these migrants receive increased and relatively stable income. Since, most of the families largely depend on remit-tance as a major source of income. Hence these migrants are highly motivated to migrate to provide a crucial finan-cial support to their families and therefore undergo the stress and strain at their work and living place.

43

20 Panna Paanch and Sattal is a type of machine used for weaving the clothes

44

Chapter 4-Discussion and Empirical Analysis on Cost of Remittance

This chapter deals with the following research objective to assess the cost, amount, frequency and channels used for remittance and its importance in rural households. At the beginning of this chapter, I provide a description as to why remittances occur, potential effects of remittance and then move on to examine money transfer mechanisms and strategies used by migrants. Thereafter, I begin to develop empirical analysis based on the primary data collected. The above stated research objective is analysed within the In-dian context.

Migrant labourers contribute upto 10% of India’s GDP (Deshingkar and Akter 2009) and the remittance they send account for as much as half their family income at migra-tion sources (Sahu and Das 2008). Although their work ac-counts for a significant contribution to economic growth, these migrant fail to gain access to formal financial sources to remit their money. In case of India, there have been few studies investigating the importance of domestic remittance and associated transaction cost (Sahu and Das 2008). Hence, in this chapter I look at strategies adopted by mi-grants to remit money, and associated cost, constraints faced by them and the significance of remittance to en-hance their livelihood situation in the source place.

Migrants not only attempt to improve their own liveli-hood situations but also send a considerable share of their earning back to their families. When remittances are per-ceived as a more stable source of income in the rural house-holds then migrants are motivated by altruism and send more money back to their families. Thus, remittance tends to have positive impact in the chain of migration process. While the underlying tone has shifted the state-of-art con-clusions are that remittance can and do have positive, neut-ral and negative implication for development (Carling 2004). The immediate effect of remittance is to enhance daily subsistence and basic needs. In this case, remittances contribute to direct poverty alleviation. A study by (Stark, Taylor and Yitzhaki 1986) shows that migrant’s remittance do not compensate for this adverse impacts because, in net terms, the remittances are either very small or go dispro-portionately to those better off. The paper also highlights risk diversification, alleviation of credit constraints and the filtering-down mechanisms pertaining to migrant remit-

45

tance, which can have some positive impact. Hence, the role of remittance depends on the magnitude of remittance in relation to income from other sources, as well as upon rankings in terms of total income of remittance in receiving households.

With this background, I first look at the need and poten-tial effect of remittance in rural households and then move on to examine the money transfer mechanism.

4.1 Need and potential effect of remittances:Remittance reflects upon the origin of the migrants them-selves as these are highly concentrated and flow to a relat-ively small number of towns and villages within the country. Remittances are qualitatively different from other sources of development finance in that they are both relatively stable and counter-cyclical in nature. The remittances from migration play a vital role in providing sustenance for the poor, and indeed, a dominant livelihood strategy (Conroy et al, 2001; Mosse et al, 2002 and Prasad, 1997).

There are two main theoretical approaches to migrant remittances. The first approach is known as “Migrant Syn-drome” (Reichart, 1981), which is inward remittances res-ulting from migration, to partially compensate for the loss of human capital. The contrarian view on migration and re-mittances is provided by the New Economics of Labour Mi-gration (NELM) (Stark and Bloom, 1985). This approach considers migration as an integral part of the household ob-jective to enhance income levels, investment capacity and acquire insurance against risk.

In this context, remittances tend to ease the production and investment constraints, mitigate risk and set in a mo-tion development dynamics. Apart from that remittance also serves as a form of insurance, as they are often countercyc-lical (Taylor, 1999). Hulme et al (2001) also argued that re-mittances are future risk premium and should enable households to accumulate assets that minimise their vulner-ability to financial shocks.

Poirine (1997) viewed remittances as an implicit family loan arrangement, exhibiting a “three waves” shape. Under this perspective, remittances are repayment made by the migrant worker to the source household for the loan con-tract made for human capital development of the migrant worker. In the first stage, migrant worker remits a signific-ant portion of income to the source family in order to repay the implicit loan obligations. During the second stage, the migrant remittances are implicit loans to siblings to finance

46

their education in the source country. According to Poirine (1997), the average aggregated value of remittances will be higher given a higher ratio of temporary migrants in the stream of total migrant, since the former passes through the three waves of migration mentioned above in a shorter period of time.

Lastly, remittances contribute to the welfare and im-prove livelihood of the receiving household – be it in terms of basic necessities such as food, clothing, or better health and education; thereby building human and social capital or to a smaller extent savings or business investments (Sander, 2003).

In conclusion, the contribution of remittance is analysed in microeconomic terms as it tends to decreases the monet-ary gap, which has risen due to lack of economic opportun-ity during natural disasters and improve the standard of liv-ing. On the other hand remittance is also invested in human and social capital like (health and education) and in build-ing assets. The remittance reduces the borrowing and eases budget constraint in the household. The aggregate demand effect generated by the remittance flow has an effect to raise economic activity in non-migrant household and lastly, the remittance flow has inter-temporal dimension that can be spread over time, implying relative stability in such flows.

In the following section I look at money transfer mech-anism adopted by migrants, constraints faced, loopholes in the system and the cost associated with each channel through which migrants choose to remit.

3.2 Money Transfer Mechanism Remittance needs have traditionally been thought to be well met by money orders from Post Offices, demand drafts by banks and through informal sources, including physical car-rying of cash by the poor themselves and their relatives and friends (Sen 2007). But in reality there is a huge gap in terms of services provided for the migrant’s workers. Therefore, in the following section I look at the difference in the services provided by financial institution (formal and in-formal) and why migrants prefer to use informal ways to re-mit money even when they charge heavy service rates.

In the case of banks and other formal financial institu-tions, they often have high entry barriers, cumbersome pro-cedure and loopholes in the system that excludes them from gaining access to these services. The financial service pro-viders (FSPs) could view it as very lucrative business. The

47

Financial service providers that cater to the poor have been drawn to the money transfer market because it offers them the opportunity to fulfill their financial goals as well as their social objectives. As a fee-based product, money transfers generate revenues and bolster FSP’s bottom line. From a social perspective, money transfers allow FSPs to deliver an additional service demanded by poor customers, at a cost potentially lower than that of mainstream providers (Isern et al 2005).

In understanding the needs of a client the FSPs should study both sending and receiving household characteristics, for example, age, socioeconomic background, and interper-sonal relationship between sender and recipient. Under-standing clients profile and patterns of seasonality of remit-tance can strongly influence them to provide other essential financial services either in the place of origin or at the des-tination place (Isern et al 2005).

This context requires a better understanding of preval-ent remittance mechanisms and a look at loopholes present in the system that puts the migrants as well as their famil-ies back home in a more vulnerable situation. The following section is being discussed within the Indian context.

With this background I look at formal Financial Institu-tions such as Banks and Non Governmental Organizations (NGO) which are hesitant to offer remittance services be-cause Reserve Bank of India (RBI) has issued Know Your Customer (KYC). Customer (In this case is a migrant) have no residential proof at their work place to open bank ac-count (Sen 2007).

Here I highlight on the cost of remittance charged by various financial service providers in formal and informal sector. The cost involved for using banking service is Rs 30/- upto an amount of Rs 10000/- and above this the bank charge is about Rs 2.50/- for every thousand and Rs 50/- for an amount if Rs 12500/-. For remittance by cash deposit, cash handling charges is extra it varies depending on size of remittance. In case of NGO the remittance charge is around 3% the cost varies depending on the size of remit-tance. Since they have high entry barriers and strong regu-latory migrants are denied access to these low cost ser-vices.

Even in odd cases where a worker has a bank account, making a draft to send money may take a whole day and at times the worker loses out on a day’s wage. These reasons exclude migrants from formal financial services.

48

The next preferred option available for these workers is the postal network to send part of their income to their fam-ilies back home – through money order. However, here too the problems are manifold. There is a time lag of 20 days, between sending money and its receipt by their families back home. Post office charges a fee of Rs. 150/- for an amount of Rs. 1000 to Rs 5000/- the cost varies depending on the size of remittance but, they charge a maximum of Rs 330/- for an amount of Rs 50000/-. Sometimes the money is misutilized by officials for money lending purposes and slowly people have lost confidence in this system.

Considering the hassle, the only option left for workers is to rely on informal sectors like the Tapawalas or private money operators. The workers in Surat receive their remu-neration in two instalments in any of the days between the 5th to 10th and 25th to 30th of each month, tapawals come to Surat around these dates for collection of remittance. Im-portantly, these tapawals come and collect money from mi-grant’s door steps and make the same available at the other end normally within a week’s time. Depending on the size of remittance, generally in multiples of Rs. 1000, the service charges vary; though the costs broadly remain within the range of Rs. 15 to Rs. 40. However, depending on the size of remittance, service charges tend to decrease with addi-tions of each unit of thousand rupees in the money transfer (Sahu and Das 2008).

Similar to the tapawalas, most private operators collect money from the door steps of the remitters. Some among them have opened counters at different location in the city where workers deposit amounts to be remitted. For a trans-fer of Rs 1000, a private operator generally charges an amount ranging from Rs 20 to Rs 40 as commission, and de-liver the amount to its destination within two or three days. For instance a maximum commission charged goes up to Rs 60 for the transfer of one thousand rupees expected to reach the payee within six hrs of receiving the deposit at the collection end (Sahu and Das 2008).

The second group, called as ‘freelance agents, collect remittance and facilitate money transfer through private operators. Since such agents collect remittances from their own set of people, the network possibly reduces the entry of other informal players into their constituencies. The com-mission collected by these freelance agents for transferring money, is the difference between the amount raised from the remittance and the rates that he pays to the private op-erators (Sahu and Das 2008).

49

The lack of formal avenues and loopholes in the system leaves the migrant workers as well as the families back home in a vulnerable condition. Therefore, these migrants opt and rely heavily on the service provided by Tapawalas, private money operator or freelance agents who also levy exploitative service charge for the meagre amount that mi-grant workers remit.

In conclusion, the framework on remittance mechanism should note the needs in the following areas:

a. Accessible service - The product should be available to them without unnecessary hassle. Entry barriers such as existing relationship with banks, no possession of checkable accounts or filling out long forms.

b. Timeliness and certainty of delivery - Predictability of delivery at the recipient's place is important for those who depend on remittances for meeting their basic needs.

c. Cost effectiveness, affordability and value for money services - The remittance needs are mostly repetitive and small value. The present system of remitting through post offices is costly. Other sources are risk prone, further adding to the cost of the service.

d. Receipt of delivery status - Timely confirmation of de-livery is a requirement for poor people who have limited ac-cess to communication facility.

3.3 Empirical Analysis: Based on the primary data set this section identifies and ex-amines the remittance patterns and strategies used by indi-viduals at the destination place to remit money back home with emphasis on the importance of remittance in the rural hh. First, I examine whether migrant send remittance, if they send the nature of remittance is addressed whether it is systematic or sporadic remittance. Thereafter, I examine the size of remittance, cost and channel used for remit-tance, followed by the cost associated for each channel and size of remittance and finally on the income from various source and the contribution of remittance to the hh income.

The above mentioned detail provides an overview on the role and importance of remittance mechanism adopted by the migrant to remit money to rural hh.

Empirical ResultsI begin my analysis by showing the types of remittance whether systematic or sporadic, and the cost associated in

50

each of case, which is then linked to channels and size of remittance. The analysis on types of frequency of remit-tance is done separately so as to provide a clear distinction in each of the case. Then, I move on to analyse the purpose of such remittance and the importance of these remittance in rural households, and lastly, the analysis of the conveni-ence of channels used and effect of remittance on the rural households.

In this section I begin to provide information from the perspective of remittance sending hh so as to analyse the cost of remittance. For the purpose of analysis I use univari-ate and bivariate tables.

Table 20: Analysis of cost, frequency and size of remittance depending on the fre-quency of remittance

Number of individual sending remit-tance

Frequency Percent

Yes 288 100.00Remittance FrequencySystematic (Yes) 265 92.01Sporadic (Yes) 118 40.97

This table provides information on remittance sending. Out of 288 individuals, all of them send remittance back to their families. Among them 92% sent systematic remit-tances. Apart from systematic remittance the individuals also send sporadic remittance which is about 41 percent. It is apparent that systematic remittance is more frequent than sporadic remittance. Even in case of systematic and sporadic remittance there are lots of variations and these are mainly attributed to the cost, channel and size of remit-tance. A study by (Samal 2006) in Andhra Pradesh reveals, that regarding the regularity of remittance 35% hh receive irregular remittance, 15% receive monthly remittance and 44% receive remittance quarterly and the rest of the remit-tance goes under advance payment. In the present study it is observed that 92.01% send remittance on a regular basis and 40.07% receive sporadic remittance. The large vari-ation in the data could be because of sample size. The present study uses 288, individuals whereas, a study done by Samal, 2006 uses only 100 sample size to collect the in-formation, the place and context of migration are different in both the cases.

Table 21: amount, channels and cost of remittance (these figures are based on per annum)

Size Frequency Percent6000-12000 26 9.03

51

14400-18000 33 11.4620000-24000 87 30.2125000-32000 45 15.6333000-44000 51 17.7145000-56000 32 11.1160000-96000 14 4.86Channels Frequency PercentBank 99 34.38Internet Banking 6 2.08Post Office 3 1.04Formal NGO agencies 165 57.29Friends/Relatives 5 1.74Others 10 3.47Cost of remittance Frequency Percent0-180 35 12.15200-420 42 14.58450-640 48 16.67645-780 34 11.81800-1000 47 16.321020-1210 25 8.681240-1510 25 8.681520-1990 21 7.292000-2880 11 3.82

In these tables I look at size, channels and cost of remit-tance incurred by these migrants at the destination place. From the above it is evident that, the size of remittance var-ies in between Rs 6000/- to Rs 96000/-. 30% migrants annu-ally remit an amount of Rs. 20000/- to Rs 24000/ followed by an amount of Rs 33000/- to Rs 44000/- which is about 18% of the total sample collected and another 16% of them remit an amount of Rs 45000/- to Rs 56000/- which is about 11%. The remaining remit either very low or large amount of remittance. In the following table I look at channels used by the migrants it is clear that about 57% of these migrant prefer to remit money through the local NGO that offers formal financial services. The next highest channels used by migrants are the Bank which is about 34%. In the case of cost of remittance it is observed that cost varies from a min-imum of Rs 180 to a maximum of Rs 2880/- and depending on the size of remittance the cost also varies. 17% of these migrants remittance cost falls in the range of Rs 450 to Rs 640, followed by 16% of them incur cost of Rs 800/- to Rs 1000/- and the rest either fall in the low or the highest range in the remittance cost.

In order to substantiate my findings I also refer to other studies done by (Sahu and Das, 2008), which reveals that 90% of migrants’ use informal source for remittance and the rest remit their money through formal sources. How-ever, there is a sharp difference in data which reveals 57.29% of the migrant use NGO service, followed by 34.38% use bank for remittance and the rest 8.33% use in-formal service for remittance. Therefore, the present study is in contrast with other studies. This difference, is manifest

52

because most of the migrants in my study used NGO ser-vice. However, from discussion with respondents and the NGO it has come out that, the present NGO is planning to with-draw the remittance service because of regulatory is-sues. This percent of people have to again rely on informal service to remit money back home and the other reason be-ing place of remittance and the presence of NGO facilities only in those areas of source place. Another study by (C.K. Samal 2006) found that seasonal and contractual labourers make more regular and substantial remittances than short-term migrants. The study also focuses on other determin-ants that affect the size of remittance such as size of hh, number of dependents and purpose of remittance.

Chart3: Size of remittance and channel used for remittance

53

Table 22: Cost of remittance with mean and plus/minus standard deviations

Channel used for remittance

Type of remittance instrument

Minimum and Maximum amount of transfer

Identity required for transferring

Cost of cash transfer

Bank Demand Draft, Telegraphic trans-fer, Money order and Bankers Cheque

Rs1000/- to Rs 10000/- Maximum amount Rs 12500/-

PAN (Permanent Account Number)

Rs 30In excess of Rs 10000/- Rs2.50/- is charged

Internet Banking Only NEFT (Na-tional Electronic Fund Transfer )

No LimitUp to Rs 100000100000-200000Above 200000Max Rs 500000

Bank account, in-ternet facility

Rs 15Rs 25

Post Office IMO (Instant Money Order) ser-vices

1000 - 50005001-1000010001-1500015001-2000020001-2500025001-3000030001-3500035001-4000040001-4500045001-50000

Voter’s Identity Card/PAN/Ration card

Rs 150Rs 170Rs 190Rs 210Rs 230Rs 250Rs 270Rs 290Rs 310Rs 330

Formal NGO agencies

Door step delivery service at both ends

Surat:Up to Rs 10001001-1000010001-1200012001-25000GandhidhamUp to Rs 500501-10001001-50005001-60006001-10000

No Identity re-quired Rs 30 +10

3% + 10Rs 300 + 102.5% + 10

30 (fixed)40 (fixed)3% + 10 1602.5 + 10

Other (Informal Agencies)

Door step delivery services at both places

No limitRs 1000For instance money transfer

No identity re-quired

Rs 15 to Rs 40. Rs 60

Source: post office India website, State bank of India website and Adhikar NGO website

54

Table 23: Cost of remittance as a proportion of amount for each channel (per month)

Channel used for remittance

Cost incurred for each channel

Size of remittance for each channel

Percent on cost of remittance as a proportion to channel

Bank 4,816 254,117 1.90Internet Banking 83 15,333 0.54Post Office 122 4,500 2.70Formal NGO agen-cies 14,104 410,225

3.44

Other (Informal Agencies) 659 25,083

3.63

Friends/ Relatives --- 6,000 ---

Table 22 and 23 provides information on cost of re-mittance. Table 22 provides information on size of remit-tance and its associate cost for each of the channels used whereas table 23 provides information on the actual data collected from the field. From these tables it is clear that there are high entry barriers in case of formal financial in-stitutions such as banks and post office because of identity proof. Even though the cost of remittance is low when com-pared to other forms of remittance, migrant prefer to use informal channels because of accessibility, door step deliv-ery service, less cumbersome and no documentation re-quired for remitting money. With these hassle in the formal sector, migrant are forced to use informal channels for re-mittance that charge heavy rates for remittance.

For example, we can look into the experience of M-PESA model adopted by Kenya, this organization provides various types services such as depositing cash to your account, transfer of money, withdrawal of money, buy safaricom sir-time, paybills and manage your M-PESA account. Our focus would be on money transfer and we do not account for other services as they are beyond the objectives of this re-search paper. But we could see apt use of technology to maximize the customer benefits.

Table 24: Information on the charges collected by M-PESA

Transaction Type Transaction Range (KShs) Customer Charge (KShs)

Minimum MaximumSend money to a registered M-PESA user

100 35000 30

Send money to non registered M-PESA users

100 2500 752501 5000 1005001 10000 17510001 20000 35020001 35000 400

Source: M-PESA website

From the above mentioned tables we can see a sharp difference not only in service charges but also on a wide

55

range of services that are provided by this organisation by mere use of technology to enhance the productivity and thereby reducing transaction cost and are able to provide better service in the most efficient, cost effective and trust-worthy manner.

In conclusion, it is noticed that there is a huge differ-ence in the service offered and also paves a way to adopt new opportunities that enables the FSP to provide cost ef-fective services to the migrants.

Table 25: Average size of remittance to cost amount in Rupees (last one year)

Information on remittance Amount PercentageAnnul average systematic remit-tance

29802.43 90.87

Annual average sporadic remit-tance

3027.43 9.22

Annual total average remittance 32829.86 100.00Average cost of remittance 823.35 3.00Monthly Average remittance 2735.82Comparison of cost of remit-tance to income

59491.00 1.38

From the above table it is observed that cost of remit-tance on average is about 3.00 per cent and the migrant re-mit about 50.00% of their earnings back to their families. When compared to the amount of remittance, the cost of re-mittance is on the high side because it is the forgone to amount by the families to receive such remittance. Hence, in aggregate 3% is quite substantial as these households mainly survive on remittance. It is observed that migrant remit 50% of his earning if he is able to access these ser-vices in a cost effective manner then the families back home are entitled to receive comparatively larger amount as a lower cost. In order to substantiate my finding I also supple-ment a study done by (Sahu and Das, 2008), which shows that on average monthly income of a migrant is Rs 3,761, an Oriya migrant worker spends around Rs 1,882 a month and the remaining amount he remits back to his family, which accounts for nearly 50 percent of his income. The study also shows that average size of remittance stands around Rs 1,427, however, this varies according to the oc-cupation. Therefore, the findings and the study are concur-rent.

In summary, I begin my analysis whether people in the destination sent remittance or not, then moved on to look at whether the remittance received are systematic or sporadic. From the analysis it was evident that more mi-grant prefer to send systematic remittance over sporadic

56

remittance the reason being that families need remittance for smoothing their consumption, to repay debts and to provide better education for their children. While, the us-age of sporadic remittance is very uneven the amount may be either low or high depending on the needs of the family. Then the analysis was on size, channels and cost of remit-tance, it shows that most of the migrant prefer to use formal NGO service to remit money and most of them remit annually around Rs 24000/-. In the following table the ana-lysis was on cost of remittance as a proportion of the amount for each channel used, it was clear that migrants remit more than 50 per cent of their earning back to their family and the associated cost was around 3 percent. The cost is high in comparison to the size of remittance. This in-fluences the pattern of remittance and the various strategies adopted by migrants while remitting money back to their families.

Table 26: Percentage of income from various sources (per annum)

Amount of income from various source

Amount Percentage

Agriculture income 12,047,50 11.79Non-agriculture income 25,740,00 25.19Total income from both activities 37,787,50 36.98Income from remittance 64,408,50 63.02Total Income 10219600 100.00

This table provides information on income from agricul-ture and non-agriculture activity. The share of income from agriculture source (11.79%) is measured to be the lowest, followed by share of income from non-agriculture activity (25.19%) and income from remittance (63.02%) which has the highest share for families that has opted to migrate.

The same study suggests that agriculture contributes around 17.8 per cent of the total household income; non-farm activity contributes around 6.5 per cent of total house-hold earning. The data from the study suggests, that as much as 42.5 per cent of the total income among these households is generated from different sources and the rest through remittance.

57

Chapter 5 - Conclusion

This paper addressed two broad aspects, internal migration and domestic remittance which are interrelated with devel-opment, finance and urban labour markets. The paper in-vestigated the cause for migration, socio-economic condi-tions of remittance sending extended households/individu-als, consequences of migration and importance of remit-tance mechanism and its role in enhancing the economic conditions at receiving end. Based on primary data collec-ted from two state of Orissa (Rural) and Gujarat (Urban) covered five districts in months of July-August 2010.

First, migration has been accounted in terms of push and pull factors and highlighted the consequence and re-turn to migration that are related to multifarious reasons. Movements have taken place especially when local econom-ies have offered limited livelihood alternatives. However, migration from the districts of Nayagad, Puri and Kurdha is characterized by the following features- distress migration and distress in migration. Hence, from the analysis it is evident that prominent reasons for migration are failure of agriculture, low productivity, unemployment opportunity and excessive debt burdens and also due to better employ-ment opportunity, industrilisation and urbanisation and bet-ter wages. Hence, migration in broader terms was treated as a livelihood coping mechanism to alleviate themselves out of poverty situation.

While examining the socio-economic conditions of households the analyses showed that over a long period mi-gration does have adverse consequences on the individual at the destination place. The working ability of migrant largely depends on the working conditions and living envir-onment. Since, most them work under strenuous conditions and lack of improper facilities has reduced their present as well as future working ability. It is evident that after certain period when migrant are unable to remit maximum money back to their families. They decide to exit from the urban la-bour market and are apparently replaced by younger mi-grant. When they decide to migrate-out most of them return to their villages with or without savings. Hence, on return these migrant tend to survive on meagre living from agri-culture and allied activities. Apart from that the migrant contribution towards the family also decreases and con-sequently results in migration of another member from that

58

family hence, migration is an inevitable process for house-holds to combat poverty at the receiving end.

Further, the analysis shows that migrants live and work in extreme difficult conditions they face a high cost and risk to migration. They are highly motivated by aspiration that makes workers to struggle for better earnings or to get hold of some job in the city else, they move to other cities like Mumbai, Vapi or Gandhidham which then leads to circular migration and puts them as well as families back home in a vulnerable situation. In the destination place they often live in harsh and hostile environment at the same time include risk of abuse and exploitation, the net benefit in some cases are low or even negative. Most of these workers live a poor life that has serious implication on their mental, emotional and physiological conditions. These factors have raised con-cern over their health that makes them even more vulner-able in urban labour market.

In addressing the issue of remittance and its importance in the hh, it was found out that migrant choice of remitting money is not driven by the cost but, are driven by access to services. In case of formal financial institutions it is found that cost of remittance is the lowest but, access these ser-vices is not friendly to migrant. In case of informal institu-tions the associated cost is higher but still migrant prefer to use these channels because of convince and flexibility that make them more attractive hence, cost is not an influential factor in determining the channels used for remittance.

Further, the study also emphasised on the importance of remittance and looked at per capita income from various sources and contribution of income which suggests that in-come from remittance accounted for around 63% of the family income. Since the sustenance of these families is largely depended on income from remittance it is important that these small amounts are channelized and made avail-able in a cost effective manner.

In conclusion, in this study, the effort was to bring out that in the process of migration and remittance a migrant is not only vulnerable, but is exploited and exposed to various kinds of risk especially related to health that has a huge im-pact on migrants present as well as future working ability (before and after migration). Hence, it is important to re-visit the structure of migration and remittance from the perspective of remittance sending households.

Word count: 19801

59

60

Appendices

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

References

Acharya, A. (2008) 'Access and Utilisation of Health Care Services in Urban Low-Income Settlements in Surat, India', Working Papers .

Ammassari, S. and R. Black (2001) Harnessing the Potential of Migration and Return to Promote Development: Applying Concepts to West Africa. International Organization for Mi-gration.

Banerjee, B. (1981) 'RURAL-URBAN MIGRATION AND FAMILY TIES: AN ANALYSIS OF FAMILY CONSIDERA-TIONS IN MIGRATION BEHAVIOUR IN INDIA*', Oxford Bulletin of Economics and Statistics 43(4): 321-355.

Bhalla, G. and P. Hazell (2003) 'Rural Employment and Poverty: Strategies to Eliminate Rural Poverty within a Gen-eration', Economic and Political Weekly 38(33): 3473-3484.

Calero, C., A.S. Bedi and R. Sparrow (2009) 'Remittances, Liquidity Constraints and Human Capital Investments in Ecuador', World Development 37(6): 1143-1154.

Carling, J. and University of Oxford. Centre on Migration, Policy and Society (2004) Policy Options for Increasing the Benefits of Remittances. Centre on Migration, Policy & So-ciety.

Chandrasekhar, C. (2007) 'The Progress of" Reform" and the Retrogression of Agriculture', Social Scientist 35(1): 61-75.

Chatterjee, C.B. (2006) Identities in Motion: Migration and Health in India. Centre for enquiry into health and allied themes.

Das, B. 'Slum Dwellers in Indian Cities: The Case of Surat in Western India', QEH Working Papers .

de Haan, A., M. Drinkwater, C. Rakodi and K. Westley (2002) Methods for Understanding Urban Poverty and Livelihoods. Livelihoods Connect.

Deshingkar, P. and S. Akter (2009) 'Migration and Human Development in India', .

76

Deshingkar, P. and D. Start (2003) Seasonal Migration for Livelihoods in India: Coping, Accumulation and Exclusion. Overseas Development Institute London.

GHATE, P. (2005) 'Serving Migrants Sustainably', Economic and Political Weekly : 1740.

Harris, J.R. and M.P. Todaro (1970) 'Migration, Unemploy-ment and Development: A Two-Sector Analysis', The Ameri-can Economic Review 60(1): 126-142.

Isern, J., R. Deshpande and J. van Doorn (2005) Crafting a Money Transfers Strategy: Guidance for Pro-Poor Financial Service Providers. Consultative group to assist the poorest (CGAP).

Maimbo, S.M. and D. Ratha (2005) Remittances: Develop-ment Impact and Future Prospects. World Bank Publica-tions.

Misra (1998) Dynamics of Rural-Urban Migration in India

Poirine, B. (1997) 'A Theory of Remittances as an Implicit Family Loan Arrangement', World Development 25(4): 589-611.

Riadh, B.E.N.J. 'RURAL-URBAN MIGRATION: ON THE HARRIS-TODARO MODEL', .

Samal, C.K. (2006) 'Remittances and Sustainable Liveli-hoods in Semi-Arid Areas', Asia Pacific Development Journal 13(2): 73-92.

Sahu and Das (2008) Income, Remittances and Urban Labour Markets

Sharma, A. and A. Bhaduri (2006) 'The “Tipping Point” in Indian Agriculture: Understanding the Withdrawal of the In-dian Rural Youth', Draft prepared for the IWMI-CPWF project on “Strategic Analysis of National River Linking Project of India .

Singh, B. (2006) 'Worker’s Remittances as Stable Financial Flows: Some Evidence from India', .

Sjaastad, L.A. (1962) 'The Costs and Returns of Human Mi-gration', The Journal of Political Economy 70(5): 80-93.

Skeldon, R. (2002) 'Migration and Poverty', Asia Pacific Population Journal 17(4): 67-82.

77

Skeldon, R. (1997) Migration and Development. Longman Harlow.

Taylor, E.J. (1999) 'The New Economics of Labour Migra-tion and the Role of Remittances in the Migration Process', International Migration 37(1): 63-88.

78